Next-generation wireless networks such as 5G and 802.11ad networks will use millimeter waves operating
at 28GHz, 38GHz, or higher frequencies to deliver unprecedentedly high data rates, e.g., 10 gigabits per
second. However, millimeter waves must be used directionally with narrow beams in order to overcome the
large attenuation due to their higher frequency. To achieve high data rates in a mobile setting,
communicating nodes need to align their beams dynamically, quickly, and in high resolution. We propose a
data-driven, deep neural network (DNN) approach to provide robust localization for beam alignment,
using a lower frequency spectrum (e.g., 2.4 GHz). The proposed DNN-based localization methods use the
angle of arrival derived from phase differences in the signal received at multiple antenna arrays to infer the
location of a mobile node. Our methods differ from others that use DNNs as a black box in that the
structure of our neural network model is tailored to address difficulties associated with the domain, such as
collinearity of the mobile node with antenna arrays, fading and multipath. We show that training our
models requires a small number of sample locations, such as 30 or fewer, making the proposed methods
practical. Our specific contributions are: (1) a structured DNN approach where the neural network
topology reflects the placement of antenna arrays, (2) a simulation platform for generating training and
evaluation data sets under multiple noise models, and (3) demonstration that our structured DNN approach
improves localization under noise by up to 25% over traditional off-the-shelf DNNs, and can achieve submeter
accuracy in a real-world experiment.
Wireless Sensor Networks are highly distributed self-organized systems. WSN have been deployed in various fields. Now a day, Topology issues have received more and more attentions in Wireless Sensor Networks (WSN). While WSN applications are normally optimized by the given underlying network topology, another trend is to optimize WSN by means of topology control. In this area, a number of approaches have been invested, like network connectivity based topology control, cooperating schemes, topology directed routing, sensor coverage based topology control. Most of the schemes have proven to be able to provide a better network monitoring and communication performance with prolonged system lifetime. In this survey paper, I provide a full view of the studies in this area.
In next five years 5G is the most popular and anticipated mobile technology and beam forming is one of the important aspect of 5G networks Beam forming is a technique used by sensor arrays for a directional signal transmission or reception and is very important for number of applications like Radar, biomedicine, radio communications, SONAR The paper introduces beam forming technique sand its importance in the modern cellular society Irfan Nissar Bhat | Er. Harish Dogra "Beamforming for 5G Networks" Published in International Journal of Trend in Scientific Research and Development (ijtsrd), ISSN: 2456-6470, Volume-2 | Issue-6 , October 2018, URL: http://www.ijtsrd.com/papers/ijtsrd18405.pdf
IGeekS Technologies is a company located in Bangalore, India. We have being recognized as a quality provider of hardware and software solutions for the student’s in order carry out their academic Projects. We offer academic projects at various academic levels ranging from graduates to masters (Diploma, BCA, BE, M. Tech, MCA, M. Sc (CS/IT)). As a part of the development training, we offer Projects in Embedded Systems & Software to the Engineering College students in all major disciplines.
Broad-Spectrum Model for Sharing Analysis between IMTAdvanced Systems and FSS...IOSRJECE
An appraisal of orthogonal frequency division multiplexing (OFDM) accredited for IMT-Advanced has been well thought-out in this letter. Derivation of the power spectral density (PSD) produce new model which easily assess the interfering signal power that appears in the band of a victim system without a spectrum emission mask. Furthermore, the broad-spectrum investigative model (BIM) can assess the interference from the 4G systems into FSS systems, when transmit power is unallocated to some sub-carriers overlapping the band of the victim system. Closed form is derived to create the model.
AN IMPROVED DECENTRALIZED APPROACH FOR TRACKING MULTIPLE MOBILE TARGETS THROU...ijwmn
Target localization and tracking problems in WSNs have received considerable attention recently, driven
by the requirement to achieve high localization accuracy, with the minimum cost possible. In WSN based
tracking applications, it is critical to know the current location of any sensor node with the minimum
energy consumed. This paper focuses on the energy consumption issue in terms of communication
between nodes whenever the localization information is transmitted to a sink node. Tracking through
WSNs can be categorized into centralized and decentralized systems. Decentralized systems offer low
power consumption when deployed to track a small number of mobile targets compared to the centralized
tracking systems. However, in several applications, it is essential to position a large number of mobile
targets. In such applications, decentralized systems offer high power consumption, since the location of
each mobile target is required to be transmitted to a sink node, and this increases the power consumption
for the whole WSN. In this paper, we propose a power efficient decentralized approach for tracking a
large number of mobile targets while offering reasonable localization accuracy through ZigBee network
Wireless Sensor Networks are highly distributed self-organized systems. WSN have been deployed in various fields. Now a day, Topology issues have received more and more attentions in Wireless Sensor Networks (WSN). While WSN applications are normally optimized by the given underlying network topology, another trend is to optimize WSN by means of topology control. In this area, a number of approaches have been invested, like network connectivity based topology control, cooperating schemes, topology directed routing, sensor coverage based topology control. Most of the schemes have proven to be able to provide a better network monitoring and communication performance with prolonged system lifetime. In this survey paper, I provide a full view of the studies in this area.
In next five years 5G is the most popular and anticipated mobile technology and beam forming is one of the important aspect of 5G networks Beam forming is a technique used by sensor arrays for a directional signal transmission or reception and is very important for number of applications like Radar, biomedicine, radio communications, SONAR The paper introduces beam forming technique sand its importance in the modern cellular society Irfan Nissar Bhat | Er. Harish Dogra "Beamforming for 5G Networks" Published in International Journal of Trend in Scientific Research and Development (ijtsrd), ISSN: 2456-6470, Volume-2 | Issue-6 , October 2018, URL: http://www.ijtsrd.com/papers/ijtsrd18405.pdf
IGeekS Technologies is a company located in Bangalore, India. We have being recognized as a quality provider of hardware and software solutions for the student’s in order carry out their academic Projects. We offer academic projects at various academic levels ranging from graduates to masters (Diploma, BCA, BE, M. Tech, MCA, M. Sc (CS/IT)). As a part of the development training, we offer Projects in Embedded Systems & Software to the Engineering College students in all major disciplines.
Broad-Spectrum Model for Sharing Analysis between IMTAdvanced Systems and FSS...IOSRJECE
An appraisal of orthogonal frequency division multiplexing (OFDM) accredited for IMT-Advanced has been well thought-out in this letter. Derivation of the power spectral density (PSD) produce new model which easily assess the interfering signal power that appears in the band of a victim system without a spectrum emission mask. Furthermore, the broad-spectrum investigative model (BIM) can assess the interference from the 4G systems into FSS systems, when transmit power is unallocated to some sub-carriers overlapping the band of the victim system. Closed form is derived to create the model.
AN IMPROVED DECENTRALIZED APPROACH FOR TRACKING MULTIPLE MOBILE TARGETS THROU...ijwmn
Target localization and tracking problems in WSNs have received considerable attention recently, driven
by the requirement to achieve high localization accuracy, with the minimum cost possible. In WSN based
tracking applications, it is critical to know the current location of any sensor node with the minimum
energy consumed. This paper focuses on the energy consumption issue in terms of communication
between nodes whenever the localization information is transmitted to a sink node. Tracking through
WSNs can be categorized into centralized and decentralized systems. Decentralized systems offer low
power consumption when deployed to track a small number of mobile targets compared to the centralized
tracking systems. However, in several applications, it is essential to position a large number of mobile
targets. In such applications, decentralized systems offer high power consumption, since the location of
each mobile target is required to be transmitted to a sink node, and this increases the power consumption
for the whole WSN. In this paper, we propose a power efficient decentralized approach for tracking a
large number of mobile targets while offering reasonable localization accuracy through ZigBee network
Energy Conservation in Wireless Sensor Networks Using Cluster-Based ApproachIJRES Journal
In a wireless networking environment, the network is comprised of sensor nodes and backbones are subsets of sensors or actuators that suffice for performing basic data communication operations. They are applied for energy efficient broadcasting. In a broadcasting (also known as data dissemination) task, a message is to be sent from one node, which could be a sink or an actuator, to all the sensors or all the actuators in the network. The goal is to minimize the number of rebroadcasts while attempting to deliver messages to all sensors or actuators. Neighbor detection and route discovery algorithms that consider a realistic physical layer are described. An adaptive broadcasting protocol without parameters suitable for delay tolerant networks is further discussed. In existing solutions for minimal energy broadcasting problem, nodes can adjust their transmission powers. Wireless Sensor Networks (WSNs) are sets of many sensors that gather data and collaborate together. So, the procedures of broadcast or multicast are more important than traditional point-to-point communication in computer network. This paper focuses on broadcasting in structured WSNs. In such a kind, the procedure of network communications is easier than in unstructured WSNs. Thus, it will make an overview of Multi Point Relay (MPR) to show its weakness. Then define a cluster-based architecture for WSNs which is constructed using MPR. Next, provide a new broadcast algorithm based on the previous cluster architecture called 3B (Backbone Based Broadcasting). By the end, an illustration of 3B shows that it minimizes the energy consumption for accomplishing broadcast compared to MPR.
Mobility and Propagation Models in Multi-hop Cognitive Radio Networksszhb
Cognitive radio networks allow unlicensed
(secondary) users to opportunistically utilize the idle
resource of a licensed network for communication
without affecting the quality of service being offered to
the primary or licensed users. This paper investigates
the effect of mobility on performance of multi-hop
cognitive radio network under various propagation
models. MPEG4 video; a bandwidth intensive traffic, is
tested over these network conditions for secondary
users and results are validated using NS2 simulations.
Performance metrics used for evaluation include
throughput, delay variations etc.
Wireless Sensor Network Based Clustering Architecture for Cooperative Communi...ijtsrd
We propose clusters based cooperatives based verbal architecture coop on the cellular ad hoc wireless sensor network Mawsn with the environment fading Rayleigh. The main ability and contributions of this paper are as follows. First, the proposed cage uses a cluster as a underlying system to help stable transmission services. 2D, the proposed enclosure uses a cluster based verbal cooperative exchange to effectively guide the package delivery ratio with multi hop power saving transmission. 0.33, we do not forget reasonable methods mainly based on cellular ad hoc nodes with sensing features and constant sensor nodes in the sensor field along with conventional research for the introduction of constant network sensors. Fourth, we have theoretical analysis with blackouts opportunities for proposed cooperative transmissions. Overall performance evaluation is run through simulation and evaluation. Sweeti Kumari | Dr. Ranjan Kumar Singh "Wireless Sensor Network Based Clustering Architecture for Cooperative Communication" Published in International Journal of Trend in Scientific Research and Development (ijtsrd), ISSN: 2456-6470, Volume-5 | Issue-4 , June 2021, URL: https://www.ijtsrd.compapers/ijtsrd43670.pdf Paper URL: https://www.ijtsrd.comengineering/electronics-and-communication-engineering/43670/wireless-sensor-network-based-clustering-architecture-for-cooperative-communication/sweeti-kumari
Characterization of directed diffusion protocol in wireless sensor networkijwmn
Wireless sensor network (WSN) has enormous applications in many places for monitoring the environments
of importance. Sensor nodes are capable of sensing, computing, and communicating. These sensor nodes
are energy constraint and operated by batteries. Since energy consumption is an important issue of WSN,
there have been many energy-efficient protocols proposed for the WSN. Directed diffusion (DD) is a datacentric
protocol that focuses on the energy efficiency of the networks. Since the first proposal of DD
protocol by Deborah, there have been various versions of DD protocols proposed by many scientists across
the globe. These upgraded versions of DD protocols add on various features to the original DD protocol
such as energy, scalability, network lifetime, security, reliability, and mobility. In this paper, we discuss
and classify various characteristics of themost populardirected diffusion protocols that have been proposed
over couple of years.
ADAPTIVE SENSOR SENSING RANGE TO MAXIMISE LIFETIME OF WIRELESS SENSOR NETWORK IJCNCJournal
Wireless Sensor Network (WSN) is commonly used to collect information from a remote area and one of the most important challenges associated with WSN is to monitor all targets in a given area while maximizing network lifetime. In wireless communication, energy consumption is proportional to the breadth of sensing range and path loss exponent. Hence, the energy consumption of communication can be minimized by varying the sensing range and decreasing the number of messages being sent. Sensing energy can be optimized by reducing the repeated coverage target. In this paper, an Adaptive Sensor Sensing Range (ASSR) technique is proposed to maximize the WSN Lifetime. This work considers a sensor network with an adaptive sensing range that are randomly deployed in the monitoring area. The sensor is adaptive in nature and can be modified in order to save power while achieving maximum time of monitoring to increase the lifetime of WSN network. The objective of ASSR is to find the best sensing range for each sensor to cover all targets in the network, which yields maximize the time of monitoring of all targets and eliminating double sensing for the same target. Experiments were conducted using an NS3 simulator to verify our proposed technique. Results show that ASSR is capable to improve the network lifetime by 20% as compared to other recent techniques in the case of a small network while achieving an 8% improvement for the case of a large networks.
Energy efficient node deployment for target coverage in wireless sensor networkGaurang Rathod
Network lifetime plays an integral role in setting up an efficient wireless sensor network. Coverage in a network needs to guarantee that the region is monitored with the required degree of reliability. Locations of sensor nodes constitute the basic input for the algorithms that examine coverage of the network. Coverage problems can be broadly classified as area coverage problem and target coverage problem. Area coverage focuses on monitoring the entire region of interest, whereas target coverage concerns monitoring only certain specific points in a given region. Target coverage can be categorized as simple coverage, k-coverage and Q-coverage.
Lower coverage level (simple coverage) is enough for environmental or habitat monitoring or applications like home security. Higher degree of coverage (k-coverage) will be required for some applications like target tracking to track the targets accurately, or if sensors work in a hostile environment such as battle fields or chemically polluted areas. More reliable results are produced for higher degree of coverage which requires multiple sensor nodes to monitor the region/targets. An example of Q-coverage is video surveillance systems deployed for monitoring hostile territorial area where some sensitive targets like a nuclear plant may need more sensors cooperate to ensure source redundancy for precise data. Sensor nodes deterministically deployed by using artificial bee colony algorithm, so as to achieve the required target coverage level and maximize the network lifetime.
A NOVEL APPROACH FOR ENERGY EFFICIENT HIERARCHY BASED ROUTING IN SENSOR NETWO...cscpconf
Wireless sensor network (WSN) is the collection of many micro-sensor nodes, connecting each other by a
wireless medium. WSN exhibits different approaches to provide reliable sensing of the environment,
detecting and reporting events. In this paper, we have proposed an algorithm for hierarchy based protocols
of wireless sensor networks, which consist of two groups of sensor nodes in a single cluster node. Each
cluster consists of a three cluster head. The event driven data sensing mechanism is used in this paper and
this sensed data is transmitted to the master section head. Hence efficient way of data transmission is possible with larger group of nodes. In this approach, using hierarchy based protocols; the lifetime of the sensor network is increased.
Scheduling different types of packets, such as
real-time and non-real-time data packets, at sensor nodes with
resource constraints in Wireless Sensor Networks (WSN) is of
vital importance to reduce sensors’ energy consumptions and endto-end
data transmission delays. Most of the existing packetscheduling
mechanisms of WSN use First Come First Served
(FCFS), non pre-emptive priority and pre-emptive priority
scheduling algorithms. These algorithms incur a high processing
overhead and long end-to-end data transmission delay due to the
FCFS concept, starvation of high priority real-time data packets
due to the transmission of a large data packet in non pre-emptive
priority scheduling, starvation of non-real-time data packets due
to the probable continuous arrival of real-time data in preemptive
priority scheduling, and improper allocation of data
packets to queues in multilevel queue scheduling algorithms.
Moreover, these algorithms are not dynamic to the changing
requirements of WSN applications since their scheduling policies
are predetermined.
In the Advanced Multilevel Priority packet scheduling
scheme, each node except those at the last level has three levels of
priority queues. According to the priority of the packet and
availability of the queue, node will schedule the packet for
transmission. Due to separated queue availability, packet
transmission delay is reduced. Due to reduction in packet
transmission delay, node can goes into sleep mode as soon as
possible. And Expired packets are deleted at the particular node
at itself before reaching the base station, so that processing
burden on the node is reduced. Thus, energy of the node is saved.
Application of Weighted Centroid Approach in Base Station Localization for Mi...IJMER
A Wireless Sensor Networks (WSNs) consisting of sensor with strategic locations, and a base-stations (BSs) whose locations are relatively flexible. A sensor cluster consists of many small sensor nodes (SNs) that capture, encode, and transmit relevant information from a designated area. This article is focused on the topology of positioning process for BSs in WSNs. Heterogeneous SNs are battery-powered and energy-constrained, their node lifetime directly affects the network lifetime of WSNs. We have proposed an algorithmic approach to locate BSs optimally such that we can maximize the topological network lifetime of WSNs deterministically, even when the initial energy provisioning for SNs is no longer always proportional to their average bit-stream rate. The obtained optimal BS locations are under different length of area field and number of nodes according to the mission criticality of WSNs. By studying energy consumption due to space loss and amplification losses in WSNs, we establish the upper and lower bounds of maximal topological parameters of area and number of nodes, which enable a quick assessment of energy provisioning feasibility and topology necessity. Numerical results and surface plot are given to demonstrate the efficiency and optimality of the proposed topology of BSs positioning approaches designed for maximizing network lifetime of WSNs.
Indoor Radio Propagation Model Analysis Wireless Node Distance and Free Space...IJERA Editor
Ultra wide bandwidth (UWB) signals are commonly defined as signals that have a large relative bandwidth
(bandwidth divided by the carrier frequency) or a large absolute bandwidth. Typical indoor environments contain
multiple walls and obstacles consisting of different materials. The RF ultra wideband (UWB) system is a
promising technology for indoor localisation owing to its high bandwidth that permits mitigation of the multipath
identification problem. The work proposed in this paper identifies exact position of transmitter and receiver
wireless nodes, calculates free space path loss and distance between two nodes by considering frequency
bandwidth using 2-point and 3-point Gaussian filter. Also in the paper three types of indoor radio propagation
models are analyzed at ultra wideband frequency range and results are compared to select best suitable model for
setting up indoor wireless connectivity and nodes in typical office, business and college environments and
WPAN applications.
International Journal of Engineering Research and Applications (IJERA) is an open access online peer reviewed international journal that publishes research and review articles in the fields of Computer Science, Neural Networks, Electrical Engineering, Software Engineering, Information Technology, Mechanical Engineering, Chemical Engineering, Plastic Engineering, Food Technology, Textile Engineering, Nano Technology & science, Power Electronics, Electronics & Communication Engineering, Computational mathematics, Image processing, Civil Engineering, Structural Engineering, Environmental Engineering, VLSI Testing & Low Power VLSI Design etc.
Beamforming with per antenna power constraint and transmit antenna selection ...sipij
In this paper, transmit beamforming and antenna selection techniques are presented for the Cooperative
Distributed Antenna System. Beamforming technique with minimum total weighted transmit power
satisfying threshold SINR and Per-Antenna Power constraints is formulated as a convex optimization
problem for the efficient performance of Distributed Antenna System (DAS). Antenna Selection technique is
implemented in this paper to select the optimum Remote Antenna Units from all the available ones. This
achieves the best compromise between capacity and system complexity. Dual polarized and Triple
Polarized systems are considered. Simulation results prove that by integrating Beamforming with DAS
enhances its performance. Also by using convex optimization in Antenna Selection enhances the
performance of multi polarized systems.
A small vessel detection using a co-located multi-frequency FMCW MIMO radar IJECEIAES
Small vessels detection is a known issue due to its low radar cross section (RCS). An existing shore-based vessel tracking radar is for long-distance commercial vessels detection. Meanwhile, a vessel-mounted radar system known for its reliability has a limitation due to its single radar coverage. The paper presented a co-located frequency modulated continuous waveform (FMCW) maritime radar for small vessel detection utilising a multiple-input multiple-output (MIMO) configuration. The radar behaviour is numerically simulated for detecting a Swerling 1 target which resembles small maritime’s vessels. The simulated MIMO configuration comprised two transmitting and receiving nodes. The proposal is to utilize a multi-frequency FMCW MIMO configuration in a maritime environment by applying the spectrum averaging (SA) to fuse MIMO received signals for range and velocity estimation. The analysis was summarised and displayed in terms of estimation error performance, probability of error and average error. The simulation outcomes an improvement of 2.2 dB for a static target, and 0.1 dB for a moving target, in resulting the 20% probability of range error with the MIMO setup. A moving vessel's effect was observed to degrade the range error estimation performance between 0.6 to 2.7 dB. Meanwhile, the proposed method was proven to improve the 20% probability of velocity error by 1.75 dB. The impact of multi-frequency MIMO was also observed to produce better average error performance.
Effective range free localization scheme for wireless sensor networkijmnct
Location aware sensors can be used in many areas such as military and civilian applications. Wireless
Sensor Networks help to identify the accurate location of the event. In this paper a cost effective schema for
localization has been proposed. It uses two beacon nodes to identify the location of unknown nodes. It
also uses flooding and estimating method to accurately identify the location of other nodes. Available area
is divided into zones and beacons are provided for each zone. Beacon nodes are placed in appropriate
locations normally two in a zone to provide location information. Using the two nodes location of unknown
nodes can be calculated accurately.
Energy Conservation in Wireless Sensor Networks Using Cluster-Based ApproachIJRES Journal
In a wireless networking environment, the network is comprised of sensor nodes and backbones are subsets of sensors or actuators that suffice for performing basic data communication operations. They are applied for energy efficient broadcasting. In a broadcasting (also known as data dissemination) task, a message is to be sent from one node, which could be a sink or an actuator, to all the sensors or all the actuators in the network. The goal is to minimize the number of rebroadcasts while attempting to deliver messages to all sensors or actuators. Neighbor detection and route discovery algorithms that consider a realistic physical layer are described. An adaptive broadcasting protocol without parameters suitable for delay tolerant networks is further discussed. In existing solutions for minimal energy broadcasting problem, nodes can adjust their transmission powers. Wireless Sensor Networks (WSNs) are sets of many sensors that gather data and collaborate together. So, the procedures of broadcast or multicast are more important than traditional point-to-point communication in computer network. This paper focuses on broadcasting in structured WSNs. In such a kind, the procedure of network communications is easier than in unstructured WSNs. Thus, it will make an overview of Multi Point Relay (MPR) to show its weakness. Then define a cluster-based architecture for WSNs which is constructed using MPR. Next, provide a new broadcast algorithm based on the previous cluster architecture called 3B (Backbone Based Broadcasting). By the end, an illustration of 3B shows that it minimizes the energy consumption for accomplishing broadcast compared to MPR.
Mobility and Propagation Models in Multi-hop Cognitive Radio Networksszhb
Cognitive radio networks allow unlicensed
(secondary) users to opportunistically utilize the idle
resource of a licensed network for communication
without affecting the quality of service being offered to
the primary or licensed users. This paper investigates
the effect of mobility on performance of multi-hop
cognitive radio network under various propagation
models. MPEG4 video; a bandwidth intensive traffic, is
tested over these network conditions for secondary
users and results are validated using NS2 simulations.
Performance metrics used for evaluation include
throughput, delay variations etc.
Wireless Sensor Network Based Clustering Architecture for Cooperative Communi...ijtsrd
We propose clusters based cooperatives based verbal architecture coop on the cellular ad hoc wireless sensor network Mawsn with the environment fading Rayleigh. The main ability and contributions of this paper are as follows. First, the proposed cage uses a cluster as a underlying system to help stable transmission services. 2D, the proposed enclosure uses a cluster based verbal cooperative exchange to effectively guide the package delivery ratio with multi hop power saving transmission. 0.33, we do not forget reasonable methods mainly based on cellular ad hoc nodes with sensing features and constant sensor nodes in the sensor field along with conventional research for the introduction of constant network sensors. Fourth, we have theoretical analysis with blackouts opportunities for proposed cooperative transmissions. Overall performance evaluation is run through simulation and evaluation. Sweeti Kumari | Dr. Ranjan Kumar Singh "Wireless Sensor Network Based Clustering Architecture for Cooperative Communication" Published in International Journal of Trend in Scientific Research and Development (ijtsrd), ISSN: 2456-6470, Volume-5 | Issue-4 , June 2021, URL: https://www.ijtsrd.compapers/ijtsrd43670.pdf Paper URL: https://www.ijtsrd.comengineering/electronics-and-communication-engineering/43670/wireless-sensor-network-based-clustering-architecture-for-cooperative-communication/sweeti-kumari
Characterization of directed diffusion protocol in wireless sensor networkijwmn
Wireless sensor network (WSN) has enormous applications in many places for monitoring the environments
of importance. Sensor nodes are capable of sensing, computing, and communicating. These sensor nodes
are energy constraint and operated by batteries. Since energy consumption is an important issue of WSN,
there have been many energy-efficient protocols proposed for the WSN. Directed diffusion (DD) is a datacentric
protocol that focuses on the energy efficiency of the networks. Since the first proposal of DD
protocol by Deborah, there have been various versions of DD protocols proposed by many scientists across
the globe. These upgraded versions of DD protocols add on various features to the original DD protocol
such as energy, scalability, network lifetime, security, reliability, and mobility. In this paper, we discuss
and classify various characteristics of themost populardirected diffusion protocols that have been proposed
over couple of years.
ADAPTIVE SENSOR SENSING RANGE TO MAXIMISE LIFETIME OF WIRELESS SENSOR NETWORK IJCNCJournal
Wireless Sensor Network (WSN) is commonly used to collect information from a remote area and one of the most important challenges associated with WSN is to monitor all targets in a given area while maximizing network lifetime. In wireless communication, energy consumption is proportional to the breadth of sensing range and path loss exponent. Hence, the energy consumption of communication can be minimized by varying the sensing range and decreasing the number of messages being sent. Sensing energy can be optimized by reducing the repeated coverage target. In this paper, an Adaptive Sensor Sensing Range (ASSR) technique is proposed to maximize the WSN Lifetime. This work considers a sensor network with an adaptive sensing range that are randomly deployed in the monitoring area. The sensor is adaptive in nature and can be modified in order to save power while achieving maximum time of monitoring to increase the lifetime of WSN network. The objective of ASSR is to find the best sensing range for each sensor to cover all targets in the network, which yields maximize the time of monitoring of all targets and eliminating double sensing for the same target. Experiments were conducted using an NS3 simulator to verify our proposed technique. Results show that ASSR is capable to improve the network lifetime by 20% as compared to other recent techniques in the case of a small network while achieving an 8% improvement for the case of a large networks.
Energy efficient node deployment for target coverage in wireless sensor networkGaurang Rathod
Network lifetime plays an integral role in setting up an efficient wireless sensor network. Coverage in a network needs to guarantee that the region is monitored with the required degree of reliability. Locations of sensor nodes constitute the basic input for the algorithms that examine coverage of the network. Coverage problems can be broadly classified as area coverage problem and target coverage problem. Area coverage focuses on monitoring the entire region of interest, whereas target coverage concerns monitoring only certain specific points in a given region. Target coverage can be categorized as simple coverage, k-coverage and Q-coverage.
Lower coverage level (simple coverage) is enough for environmental or habitat monitoring or applications like home security. Higher degree of coverage (k-coverage) will be required for some applications like target tracking to track the targets accurately, or if sensors work in a hostile environment such as battle fields or chemically polluted areas. More reliable results are produced for higher degree of coverage which requires multiple sensor nodes to monitor the region/targets. An example of Q-coverage is video surveillance systems deployed for monitoring hostile territorial area where some sensitive targets like a nuclear plant may need more sensors cooperate to ensure source redundancy for precise data. Sensor nodes deterministically deployed by using artificial bee colony algorithm, so as to achieve the required target coverage level and maximize the network lifetime.
A NOVEL APPROACH FOR ENERGY EFFICIENT HIERARCHY BASED ROUTING IN SENSOR NETWO...cscpconf
Wireless sensor network (WSN) is the collection of many micro-sensor nodes, connecting each other by a
wireless medium. WSN exhibits different approaches to provide reliable sensing of the environment,
detecting and reporting events. In this paper, we have proposed an algorithm for hierarchy based protocols
of wireless sensor networks, which consist of two groups of sensor nodes in a single cluster node. Each
cluster consists of a three cluster head. The event driven data sensing mechanism is used in this paper and
this sensed data is transmitted to the master section head. Hence efficient way of data transmission is possible with larger group of nodes. In this approach, using hierarchy based protocols; the lifetime of the sensor network is increased.
Scheduling different types of packets, such as
real-time and non-real-time data packets, at sensor nodes with
resource constraints in Wireless Sensor Networks (WSN) is of
vital importance to reduce sensors’ energy consumptions and endto-end
data transmission delays. Most of the existing packetscheduling
mechanisms of WSN use First Come First Served
(FCFS), non pre-emptive priority and pre-emptive priority
scheduling algorithms. These algorithms incur a high processing
overhead and long end-to-end data transmission delay due to the
FCFS concept, starvation of high priority real-time data packets
due to the transmission of a large data packet in non pre-emptive
priority scheduling, starvation of non-real-time data packets due
to the probable continuous arrival of real-time data in preemptive
priority scheduling, and improper allocation of data
packets to queues in multilevel queue scheduling algorithms.
Moreover, these algorithms are not dynamic to the changing
requirements of WSN applications since their scheduling policies
are predetermined.
In the Advanced Multilevel Priority packet scheduling
scheme, each node except those at the last level has three levels of
priority queues. According to the priority of the packet and
availability of the queue, node will schedule the packet for
transmission. Due to separated queue availability, packet
transmission delay is reduced. Due to reduction in packet
transmission delay, node can goes into sleep mode as soon as
possible. And Expired packets are deleted at the particular node
at itself before reaching the base station, so that processing
burden on the node is reduced. Thus, energy of the node is saved.
Application of Weighted Centroid Approach in Base Station Localization for Mi...IJMER
A Wireless Sensor Networks (WSNs) consisting of sensor with strategic locations, and a base-stations (BSs) whose locations are relatively flexible. A sensor cluster consists of many small sensor nodes (SNs) that capture, encode, and transmit relevant information from a designated area. This article is focused on the topology of positioning process for BSs in WSNs. Heterogeneous SNs are battery-powered and energy-constrained, their node lifetime directly affects the network lifetime of WSNs. We have proposed an algorithmic approach to locate BSs optimally such that we can maximize the topological network lifetime of WSNs deterministically, even when the initial energy provisioning for SNs is no longer always proportional to their average bit-stream rate. The obtained optimal BS locations are under different length of area field and number of nodes according to the mission criticality of WSNs. By studying energy consumption due to space loss and amplification losses in WSNs, we establish the upper and lower bounds of maximal topological parameters of area and number of nodes, which enable a quick assessment of energy provisioning feasibility and topology necessity. Numerical results and surface plot are given to demonstrate the efficiency and optimality of the proposed topology of BSs positioning approaches designed for maximizing network lifetime of WSNs.
Indoor Radio Propagation Model Analysis Wireless Node Distance and Free Space...IJERA Editor
Ultra wide bandwidth (UWB) signals are commonly defined as signals that have a large relative bandwidth
(bandwidth divided by the carrier frequency) or a large absolute bandwidth. Typical indoor environments contain
multiple walls and obstacles consisting of different materials. The RF ultra wideband (UWB) system is a
promising technology for indoor localisation owing to its high bandwidth that permits mitigation of the multipath
identification problem. The work proposed in this paper identifies exact position of transmitter and receiver
wireless nodes, calculates free space path loss and distance between two nodes by considering frequency
bandwidth using 2-point and 3-point Gaussian filter. Also in the paper three types of indoor radio propagation
models are analyzed at ultra wideband frequency range and results are compared to select best suitable model for
setting up indoor wireless connectivity and nodes in typical office, business and college environments and
WPAN applications.
International Journal of Engineering Research and Applications (IJERA) is an open access online peer reviewed international journal that publishes research and review articles in the fields of Computer Science, Neural Networks, Electrical Engineering, Software Engineering, Information Technology, Mechanical Engineering, Chemical Engineering, Plastic Engineering, Food Technology, Textile Engineering, Nano Technology & science, Power Electronics, Electronics & Communication Engineering, Computational mathematics, Image processing, Civil Engineering, Structural Engineering, Environmental Engineering, VLSI Testing & Low Power VLSI Design etc.
Beamforming with per antenna power constraint and transmit antenna selection ...sipij
In this paper, transmit beamforming and antenna selection techniques are presented for the Cooperative
Distributed Antenna System. Beamforming technique with minimum total weighted transmit power
satisfying threshold SINR and Per-Antenna Power constraints is formulated as a convex optimization
problem for the efficient performance of Distributed Antenna System (DAS). Antenna Selection technique is
implemented in this paper to select the optimum Remote Antenna Units from all the available ones. This
achieves the best compromise between capacity and system complexity. Dual polarized and Triple
Polarized systems are considered. Simulation results prove that by integrating Beamforming with DAS
enhances its performance. Also by using convex optimization in Antenna Selection enhances the
performance of multi polarized systems.
A small vessel detection using a co-located multi-frequency FMCW MIMO radar IJECEIAES
Small vessels detection is a known issue due to its low radar cross section (RCS). An existing shore-based vessel tracking radar is for long-distance commercial vessels detection. Meanwhile, a vessel-mounted radar system known for its reliability has a limitation due to its single radar coverage. The paper presented a co-located frequency modulated continuous waveform (FMCW) maritime radar for small vessel detection utilising a multiple-input multiple-output (MIMO) configuration. The radar behaviour is numerically simulated for detecting a Swerling 1 target which resembles small maritime’s vessels. The simulated MIMO configuration comprised two transmitting and receiving nodes. The proposal is to utilize a multi-frequency FMCW MIMO configuration in a maritime environment by applying the spectrum averaging (SA) to fuse MIMO received signals for range and velocity estimation. The analysis was summarised and displayed in terms of estimation error performance, probability of error and average error. The simulation outcomes an improvement of 2.2 dB for a static target, and 0.1 dB for a moving target, in resulting the 20% probability of range error with the MIMO setup. A moving vessel's effect was observed to degrade the range error estimation performance between 0.6 to 2.7 dB. Meanwhile, the proposed method was proven to improve the 20% probability of velocity error by 1.75 dB. The impact of multi-frequency MIMO was also observed to produce better average error performance.
Effective range free localization scheme for wireless sensor networkijmnct
Location aware sensors can be used in many areas such as military and civilian applications. Wireless
Sensor Networks help to identify the accurate location of the event. In this paper a cost effective schema for
localization has been proposed. It uses two beacon nodes to identify the location of unknown nodes. It
also uses flooding and estimating method to accurately identify the location of other nodes. Available area
is divided into zones and beacons are provided for each zone. Beacon nodes are placed in appropriate
locations normally two in a zone to provide location information. Using the two nodes location of unknown
nodes can be calculated accurately.
DATA TRANSMISSION IN WIRELESS SENSOR NETWORKS FOR EFFECTIVE AND SECURE COMMUN...IJEEE
Data transmission occurs from transmitting node to sink node, which communicate each other via large number of intermediate nodes or directly to an external base station. A network consists of numbers of nodes with one as a source and one or more as a destination node.
EFFECTIVE AND SECURE DATA COMMUNICATION IN WSNs CONSIDERING TRANSFER MODULE O...IJEEE
A Bio-inspired clustering algorithm based on BFO has been proposed and investigation on energy efficient clustering algorithms related to WSNs has been done in this paper. The contribution of this paper related to use of Bacteria foraging algorithm firstly for WSNs for enhancing network lifetime of sensor nodes.
Impact of Next Generation Cognitive Radio Network on the Wireless Green Eco s...ijeei-iaes
Land mobile communication is burdened with typical propagation constraints due to the channel characteristics in radio systems.Also,the propagation characteristics vary form place to place and also as the mobile unit moves,from time to time.Hence,the tramsmission path between transmitter and receiver varies from simple direct LOS to the one which is severely obstructed by buildings, foliage and terrain. Multipath propagation and shadow fading effects affect the signal strength of an arbitrary Transmitter-Receiver due to the rapid fluctuations in the phase and amplitude of signal which also determines the average power over an area of tens or hundreds of meters. Shadowing introduces additional fluctuations, so the received local mean power varies around the area –mean. The present paper deals with the performance analysis of impact of next generation wireless cognitive radio network on wireless green eco system through signal and interference level based k coverage probability under the shadow fading effects.
LABORATORY ANALYSIS ON THE PERFORMANCE OF 5G NSA COMMUNICATION IN A SUBURBAN ...ijwmn
The propagation of information by electromagnetic waves suffers different types of interference, according
to the characteristics of the environment. The 5G system relies on adaptive modulation and coding
techniques to better suit the channel and maximize effective data exchange between the user equipment and
the network. Practical studies on the behaviour of the system under different environmental conditions,
subject to attenuation processes such as fading, are important to understand and improve the 5G
efficiency. This work has analysed the effect of the MCS (Modulation and Coding Scheme) variation on
throughput for channel degraded by the multipath fading effect in a mobile communication. The analysis
was carried out showing that the decision algorithms in terms of MCS switching to maintain adequate data
rates according to the requirement (QoS) is an important factor. Considering both 64 QAM and 256 QAM,
the throughput degradation effect was more evident in higher-order modulations due to the higher
probability of bit error in the symbol constellation. This study can be a key for understanding and
developing robust MCS switcher for 5G and beyond communications.
LABORATORY ANALYSIS ON THE PERFORMANCE OF 5G NSA COMMUNICATION IN A SUBURBAN ...ijwmn
The propagation of information by electromagnetic waves suffers different types of interference, according
to the characteristics of the environment. The 5G system relies on adaptive modulation and coding
techniques to better suit the channel and maximize effective data exchange between the user equipment and
the network. Practical studies on the behaviour of the system under different environmental conditions,
subject to attenuation processes such as fading, are important to understand and improve the 5G
efficiency. This work has analysed the effect of the MCS (Modulation and Coding Scheme) variation on
throughput for channel degraded by the multipath fading effect in a mobile communication. The analysis
was carried out showing that the decision algorithms in terms of MCS switching to maintain adequate data
rates according to the requirement (QoS) is an important factor. Considering both 64 QAM and 256 QAM,
the throughput degradation effect was more evident in higher-order modulations due to the higher
probability of bit error in the symbol constellation. This study can be a key for understanding and
developing robust MCS switcher for 5G and beyond communications.
International Journal of Engineering and Science Invention (IJESI)inventionjournals
International Journal of Engineering and Science Invention (IJESI) is an international journal intended for professionals and researchers in all fields of computer science and electronics. IJESI publishes research articles and reviews within the whole field Engineering Science and Technology, new teaching methods, assessment, validation and the impact of new technologies and it will continue to provide information on the latest trends and developments in this ever-expanding subject. The publications of papers are selected through double peer reviewed to ensure originality, relevance, and readability. The articles published in our journal can be accessed online.
A wireless network consists of a set of wireless nodes forming the network. The bandwidth allocation scheme used in wireless networks should automatically adapt to the network’s environments, where issues such as mobility are highly variable. This paper proposes a method to distribute the bandwidth for wireless network nodes depending on dynamic methodology;this methodology uses intelligent clustering techniques that depend on the student’s distribution at the university campus, rather than the classical allocation methods. We propose a clustering-based approach to solve the dynamic bandwidth allocation problem in wireless networks, enabling wireless nodes to adapt their bandwidth allocation according to the changing number of expected users over time. The proposed solution allows the optimal online bandwidth allocation based on the data extracted from the lectures timetable, and fed to the wireless network control nodes, allowing them to adapt to their environment. The environment data is processed and clustered using the KMeans clustering algorithm to identify potential peak times for every wireless node. The proposed solution feasibility is tested by applying the approach to a case study, at the Arab American University campus wireless network.
Novel Position Estimation using Differential Timing Information for Asynchron...IJCNCJournal
Positioning techniques have been a common objective since the early development of wireless networks. However, current positioning methods in cellular networks, for instance, are still primarily focused on the use of the Global Navigation Satellite System (GNSS), which has several limitations, like high power drainage and failure in indoor scenarios. This study introduces a novel approach employing standard LTE signaling in order to provide high accuracy positioning estimation. The proposed technique is designed in analogy to the human sound localization system, eliminating the need of having information from three spatially diverse Base Stations (BSs). This is inspired by the perfect human 3D sound localization with two ears. A field study is carried out in a dense urban city to verify the accuracy of the proposed technique, with more than 20 thousand measurement samples collected. The achieved positioning accuracy is meeting the latest Federal Communications Commission (FCC) requirements in the planner dimension.
NOVEL POSITION ESTIMATION USING DIFFERENTIAL TIMING INFORMATION FOR ASYNCHRON...IJCNCJournal
Positioning techniques have been a common objective since the early development of wireless networks. However, current positioning methods in cellular networks, for instance, are still primarily focused on the use of the Global Navigation Satellite System (GNSS), which has several limitations, like high power drainage and failure in indoor scenarios. This study introduces a novel approach employing standard LTE signaling in order to provide high accuracy positioning estimation. The proposed technique is designed in analogy to the human sound localization system, eliminating the need of having information from three spatially diverse Base Stations (BSs). This is inspired by the perfect human 3D sound localization with two ears. A field study is carried out in a dense urban city to verify the accuracy of the proposed technique, with more than 20 thousand measurement samples collected. The achieved positioning accuracy is meeting the latest Federal Communications Commission (FCC) requirements in the planner dimension.
INTERFERENCE-AWARE CHANNEL ASSIGNMENT FOR MAXIMIZING THROUGHPUT IN WMN pijans
Wireless Mesh network (WMN) is dynamically self-organizing and self-configured, with the nodes in the
network automatically establishing an ad-hoc network and maintaining the mesh connectivity. The ability
to use multiple-radios and multiple channels can be cashed to increase aggregate throughput of wireless
mesh network. Thus the efficient use of available interfaces and channels without interference becomes
the key factor. In this paper we propose interference aware clustered based channel assignment schemes
which minimizes the interference and increases throughput. In our proposed scheme we have given
priority to minimize interference from nearby mesh nodes in interference range than maximizing channel
diversity. We simulated our proposed work using NS-3 and results show that our scheme improves
network performance than BFSCA and Distributed Greedy CA.
Localization is one of the key technologies in wireless sensor networks (WSNs), since it provides
fundamental support for many location-aware protocols and applications. Constraints on cost and power
consumption make it infeasible to equip each sensor node in the network with a global position system
(GPS) unit, especially for large-scale WSNs. A promising method to localize unknown nodes is to use
anchor nodes, which are equipped with GPS units among unknown nodes and broadcast their current
locations to help nearby unknown nodes with localization. In this paper we can proposed a novel algorithm
of cuboid localization with the help of central point precision method. Simulation shows that the results are
far better then existing cuboid methods and gain accuracy of up to 83% with a localization error of 1.6m
and standard deviation of 2.7.
Adaptive Sensor Sensing Range to Maximise Lifetime of Wireless Sensor NetworkIJCNCJournal
Wireless Sensor Network (WSN) is commonly used to collect information from a remote area and one of the most important challenges associated with WSN is to monitor all targets in a given area while maximizing network lifetime. In wireless communication, energy consumption is proportional to the breadth of sensing range and path loss exponent. Hence, the energy consumption of communication can be minimized by varying the sensing range and decreasing the number of messages being sent. Sensing energy can be optimized by reducing the repeated coverage target. In this paper, an Adaptive Sensor Sensing Range (ASSR) technique is proposed to maximize the WSN Lifetime. This work considers a sensor network with an adaptive sensing range that are randomly deployed in the monitoring area. The sensor is adaptive in nature and can be modified in order to save power while achieving maximum time of monitoring to increase the lifetime of WSN network. The objective of ASSR is to find the best sensing range for each sensor to cover all targets in the network, which yields maximize the time of monitoring of all targets and eliminating double sensing for the same target. Experiments were conducted using an NS3 simulator to verify our proposed technique. Results show that ASSR is capable to improve the network lifetime by 20% as compared to other recent techniques in the case of a small network while achieving an 8% improvement for the case of a large networks.
PERFORMANCE STUDY AND SIMULATION OF AN ANYCAST PROTOCOL FOR WIRELESS MOBILE A...ijwmn
This paper conducts a detailed simulation study of stateless anycast routing in a mobile wireless ad hoc
network. The model covers all the fundamental aspects of such networks with a routing mechanism using
a scheme of orientation-dependent inter-node communication links. The simulation system Winsim is used
which explicitly represents parallelism of events and processes in the network. The purpose of these
simulations is to investigate the effect of node’s maximum speed, and different TTL over the network
performance under two different scenarios. Simulation study investigates five practically important
performance metrics of a wireless mobile ad hoc network and shows the dependence of this metrics on
the transmission radius, link availability, and maximal possible node speed
AN EFFICIENT DEPLOYMENT APPROACH FOR IMPROVED COVERAGE IN WIRELESS SENSOR NET...csandit
Wireless Sensor Networks (WSNs) are experiencing a revival of interest and a continuous advancement in various scientific and industrial fields. WSNs offer favorable low cost and readily deployable solutions to perform the monitoring, target tracking, and recognition of physical events. The foremost step required for these types of ad-hoc networks is to deploy all the sensor nodes in their positions carefully to form an efficient network. Such network should satisfy the quality of service (QoS) requirements in order to achieve high performance levels. In
this paper we address the coverage requirement and its relation with WSN nodes placement problems. In fact, we present a new optimization approach based on the Flower Pollination Algorithm (FPA) to find the best placement topologies in terms of coverage maximization. We have compared the performance of the resulting algorithm, called FPACO, with the original practical swarm optimization (PSO) and the genetic algorithm (GA). In all the test instances, FPACO performs better than all other algorithms.
Similar to A STRUCTURED DEEP NEURAL NETWORK FOR DATA-DRIVEN LOCALIZATION IN HIGH FREQUENCY WIRELESS NETWORKS (20)
Vehicle Ad Hoc Networks (VANETs) have become a viable technology to improve traffic flow and safety on the roads. Due to its effectiveness and scalability, the Wingsuit Search-based Optimised Link State Routing Protocol (WS-OLSR) is frequently used for data distribution in VANETs. However, the selection of MultiPoint Relays (MPRs) plays a pivotal role in WS-OLSR's performance. This paper presents an improved MPR selection algorithm tailored to WS-OLSR, designed to enhance the overall routing efficiency and reduce overhead. The analysis found that the current OLSR protocol has problems such as redundancy of HELLO and TC message packets or failure to update routing information in time, so a WS-OLSR routing protocol based on improved-MPR selection algorithm was proposed. Firstly, factors such as node mobility and link changes are comprehensively considered to reflect network topology changes, and the broadcast cycle of node HELLO messages is controlled through topology changes. Secondly, a new MPR selection algorithm is proposed, considering link stability issues and nodes. Finally, evaluate its effectiveness in terms of packet delivery ratio, end-to-end delay, and control message overhead. Simulation results demonstrate the superior performance of our improved MR selection algorithm when compared to traditional approaches.
A Novel Medium Access Control Strategy for Heterogeneous Traffic in Wireless ...IJCNCJournal
So far, Wireless Body Area Networks (WBANs) have played a pivotal role in driving the development of intelligent healthcare systems with broad applicability across various domains. Each WBAN consists of one or more types of sensors that can be embedded in clothing, attached directly to the body, or even implanted beneath an individual's skin. These sensors typically serve asingle application. However, the traffic generated by each sensor may have distinct requirements. This diversity necessitates a dual approach: tailored treatment based on the specific needs of each traffic typeand the fulfillment of application requirements, such asreliability and timeliness. Never the less, the presence of energy constraints and the unreliable nature of wireless communications make QoS provisioning under such networks a non-trivial task. In this context, the current paper introduces a novel Medium AccessControl (MAC) strategy for the regular traffic applications of WBANs, designed to significantly enhance efficiency when compared to the established MAC protocols IEEE 802.15.4 and IEEE 802.15.6, with a particular focus on improving reliability, timeliness, and energy efficiency.
May_2024 Top 10 Read Articles in Computer Networks & Communications.pdfIJCNCJournal
The International Journal of Computer Networks & Communications (IJCNC) is a bi monthly open access peer-reviewed journal that publishes articles which contribute new results in all areas of Computer Networks & Communications. The journal focuses on all technical and practical aspects of Computer Networks & data Communications. The goal of this journal is to bring together researchers and practitioners from academia and industry to focus on advanced networking concepts and establishing new collaborations in these areas.
A Topology Control Algorithm Taking into Account Energy and Quality of Transm...IJCNCJournal
The efficient use of energy in wireless sensor networks is critical for extending node lifetime. The network topology is one of the factors that have a significant impact on the energy usage at the nodes and the quality of transmission (QoT) in the network. We propose a topology control algorithm for software-defined wireless sensor networks (SDWSNs) in this paper. Our method is to formulate topology control algorithm as a nonlinear programming (NP) problem with the objective to optimizing two metrics, maximum communication range, and desired degree. This NP problem is solved at the SDWSN controller by employing the genetic algorithm (GA) to determine the best topology. The simulation results show that the proposed algorithm outperforms the MaxPower algorithm in terms of average node degree and energy expansion ratio.
Multi-Server user Authentication Scheme for Privacy Preservation with Fuzzy C...IJCNCJournal
The integration of artificial intelligence technology with a scalable Internet of Things (IoT) platform facilitates diverse smart communication services, allowing remote users to access services from anywhere at any time. The multi-server environment within IoT introduces a flexible security service model, enabling users to interact with any server through a single registration. To ensure secure and privacy preservation services for resources, an authentication scheme is essential. Zhao et al. recently introduced a user authentication scheme for the multi-server environment, utilizing passwords and smart cards, claiming resilience against well-known attacks. This paper conducts cryptanalysis on Zhao et al.'s scheme, focusing on denial of service and privacy attacks, revealing a lack of user-friendliness. Subsequently, we propose a new multi-server user authentication scheme for privacy preservation with fuzzy commitment over the IoT environment, addressing the shortcomings of Zhao et al.'s scheme. Formal security verification of the proposed scheme is conducted using the ProVerif simulation tool. Through both formal and informal security analyses, we demonstrate that the proposed scheme is resilient against various known attacks and those identified in Zhao et al.'s scheme.
Advanced Privacy Scheme to Improve Road Safety in Smart Transportation SystemsIJCNCJournal
In -Vehicle Ad-Hoc Network (VANET), vehicles continuously transmit and receive spatiotemporal data with neighboring vehicles, thereby establishing a comprehensive 360-degree traffic awareness system. Vehicular Network safety applications facilitate the transmission of messages between vehicles that are near each other, at regular intervals, enhancing drivers' contextual understanding of the driving environment and significantly improving traffic safety. Privacy schemes in VANETs are vital to safeguard vehicles’ identities and their associated owners or drivers. Privacy schemes prevent unauthorized parties from linking the vehicle's communications to a specific real-world identity by employing techniques such as pseudonyms, randomization, or cryptographic protocols. Nevertheless, these communications frequently contain important vehicle information that malevolent groups could use to Monitor the vehicle over a long period. The acquisition of this shared data has the potential to facilitate the reconstruction of vehicle trajectories, thereby posing a potential risk to the privacy of the driver. Addressing the critical challenge of developing effective and scalable privacy-preserving protocols for communication in vehicle networks is of the highest priority. These protocols aim to reduce the transmission of confidential data while ensuring the required level of communication. This paper aims to propose an Advanced Privacy Vehicle Scheme (APV) that periodically changes pseudonyms to protect vehicle identities and improve privacy. The APV scheme utilizes a concept called the silent period, which involves changing the pseudonym of a vehicle periodically based on the tracking of neighboring vehicles. The pseudonym is a temporary identifier that vehicles use to communicate with each other in a VANET. By changing the pseudonym regularly, the APV scheme makes it difficult for unauthorized entities to link a vehicle's communications to its real-world identity. The proposed APV is compared to the SLOW, RSP, CAPS, and CPN techniques. The data indicates that the efficiency of APV is a better improvement in privacy metrics. It is evident that the AVP offers enhanced safety for vehicles during transportation in the smart city.
April 2024 - Top 10 Read Articles in Computer Networks & CommunicationsIJCNCJournal
The International Journal of Computer Networks & Communications (IJCNC) is a bi monthly open access peer-reviewed journal that publishes articles which contribute new results in all areas of Computer Networks & Communications. The journal focuses on all technical and practical aspects of Computer Networks & data Communications. The goal of this journal is to bring together researchers and practitioners from academia and industry to focus on advanced networking concepts and establishing new collaborations in these areas.
DEF: Deep Ensemble Neural Network Classifier for Android Malware DetectionIJCNCJournal
Malware is one of the threats to security of computer networks and information systems. Since malware instances are available sufficiently, there is increased interest among researchers on usage of Artificial Intelligence (AI). Of late AI-enabled methods such as machine learning (ML) and deep learning paved way for solving many real-world problems. As it is a learning-based approach, accumulated training samples help in improving thequality of training and thus leveraging malware detection accuracy. Existing deep learning methods are focusing on learning-based malware detection systems. However, there is need for improving the state of the art through ensemble approach. Towards this end, in this paper we proposed a framework known as Deep Ensemble Framework (DEF) for automatic malware detection. The framework obtains features from training samples. From given malware instance a grayscale image is generated. There is another process to extract the opcode sequences. Convolutional Neural Network (CNN) and Long Short Term Memory (LSTM) techniques are used to obtain grayscale image and opcode sequence respectively. Afterwards, a stacking ensemble is employed in order to achieve efficient malware detection and classification. Malware samples collected fromthe Internet sources and Microsoft are used for theempirical study. An algorithm known as Ensemble Learning for Automatic Malware Detection (EL-AML) is proposed to realize our framework. Another algorithm named Pre-Process is proposed to assist the EL-AML algorithm for obtaining intermediate features required by CNN and LSTM.Empirical study reveals that our framework outperforms many existing methods in terms of speed-up and accuracy.
High Performance NMF Based Intrusion Detection System for Big Data IOT TrafficIJCNCJournal
With the emergence of smart devices and the Internet of Things (IoT), millions of users connected to the network produce massive network traffic datasets. These vast datasets of network traffic, Big Data are challenging to store, deal with and analyse using a single computer. In this paper we developed parallel implementation using a High Performance Computer (HPC) for the Non-Negative Matrix Factorization technique as an engine for an Intrusion Detection System (HPC-NMF-IDS). The large IoT traffic datasets of order of millions samples are distributed evenly on all the computing cores for both storage and speedup purpose. The distribution of computing tasks involved in the Matrix Factorization takes into account the reduction of the communication cost between the computing cores. The experiments we conducted on the proposed HPC-IDS-NMF give better results than the traditional ML-based intrusion detection systems. We could train the HPC model with datasets of one million samples in only 31 seconds instead of the 40 minutes using one processor), that is a speed up of 87 times. Moreover, we have got an excellent detection accuracy rate of 98% for KDD dataset.
A Novel Medium Access Control Strategy for Heterogeneous Traffic in Wireless ...IJCNCJournal
So far, Wireless Body Area Networks (WBANs) have played a pivotal role in driving the development of intelligent healthcare systems with broad applicability across various domains. Each WBAN consists of one or more types of sensors that can be embedded in clothing, attached directly to the body, or even implanted beneath an individual's skin. These sensors typically serve asingle application. However, the traffic generated by each sensor may have distinct requirements. This diversity necessitates a dual approach: tailored treatment based on the specific needs of each traffic typeand the fulfillment of application requirements, such asreliability and timeliness. Never the less, the presence of energy constraints and the unreliable nature of wireless communications make QoS provisioning under such networks a non-trivial task. In this context, the current paper introduces a novel Medium AccessControl (MAC) strategy for the regular traffic applications of WBANs, designed to significantly enhance efficiency when compared to the established MAC protocols IEEE 802.15.4 and IEEE 802.15.6, with a particular focus on improving reliability, timeliness, and energy efficiency.
A Topology Control Algorithm Taking into Account Energy and Quality of Transm...IJCNCJournal
The efficient use of energy in wireless sensor networks is critical for extending node lifetime. The network topology is one of the factors that have a significant impact on the energy usage at the nodes and the quality of transmission (QoT) in the network. We propose a topology control algorithm for software-defined wireless sensor networks (SDWSNs) in this paper. Our method is to formulate topology control algorithm as a nonlinear programming (NP) problem with the objective to optimizing two metrics, maximum communication range, and desired degree. This NP problem is solved at the SDWSN controller by employing the genetic algorithm (GA) to determine the best topology. The simulation results show that the proposed algorithm outperforms the MaxPower algorithm in terms of average node degree and energy expansion ratio.
Multi-Server user Authentication Scheme for Privacy Preservation with Fuzzy C...IJCNCJournal
The integration of artificial intelligence technology with a scalable Internet of Things (IoT) platform facilitates diverse smart communication services, allowing remote users to access services from anywhere at any time. The multi-server environment within IoT introduces a flexible security service model, enabling users to interact with any server through a single registration. To ensure secure and privacy preservation services for resources, an authentication scheme is essential. Zhao et al. recently introduced a user authentication scheme for the multi-server environment, utilizing passwords and smart cards, claiming resilience against well-known attacks. This paper conducts cryptanalysis on Zhao et al.'s scheme, focusing on denial of service and privacy attacks, revealing a lack of user-friendliness. Subsequently, we propose a new multi-server user authentication scheme for privacy preservation with fuzzy commitment over the IoT environment, addressing the shortcomings of Zhao et al.'s scheme. Formal security verification of the proposed scheme is conducted using the ProVerif simulation tool. Through both formal and informal security analyses, we demonstrate that the proposed scheme is resilient against various known attacks and those identified in Zhao et al.'s scheme.
Advanced Privacy Scheme to Improve Road Safety in Smart Transportation SystemsIJCNCJournal
In -Vehicle Ad-Hoc Network (VANET), vehicles continuously transmit and receive spatiotemporal data with neighboring vehicles, thereby establishing a comprehensive 360-degree traffic awareness system. Vehicular Network safety applications facilitate the transmission of messages between vehicles that are near each other, at regular intervals, enhancing drivers' contextual understanding of the driving environment and significantly improving traffic safety. Privacy schemes in VANETs are vital to safeguard vehicles’ identities and their associated owners or drivers. Privacy schemes prevent unauthorized parties from linking the vehicle's communications to a specific real-world identity by employing techniques such as pseudonyms, randomization, or cryptographic protocols. Nevertheless, these communications frequently contain important vehicle information that malevolent groups could use to Monitor the vehicle over a long period. The acquisition of this shared data has the potential to facilitate the reconstruction of vehicle trajectories, thereby posing a potential risk to the privacy of the driver. Addressing the critical challenge of developing effective and scalable privacy-preserving protocols for communication in vehicle networks is of the highest priority. These protocols aim to reduce the transmission of confidential data while ensuring the required level of communication. This paper aims to propose an Advanced Privacy Vehicle Scheme (APV) that periodically changes pseudonyms to protect vehicle identities and improve privacy. The APV scheme utilizes a concept called the silent period, which involves changing the pseudonym of a vehicle periodically based on the tracking of neighboring vehicles. The pseudonym is a temporary identifier that vehicles use to communicate with each other in a VANET. By changing the pseudonym regularly, the APV scheme makes it difficult for unauthorized entities to link a vehicle's communications to its real-world identity. The proposed APV is compared to the SLOW, RSP, CAPS, and CPN techniques. The data indicates that the efficiency of APV is a better improvement in privacy metrics. It is evident that the AVP offers enhanced safety for vehicles during transportation in the smart city.
DEF: Deep Ensemble Neural Network Classifier for Android Malware DetectionIJCNCJournal
Malware is one of the threats to security of computer networks and information systems. Since malware instances are available sufficiently, there is increased interest among researchers on usage of Artificial Intelligence (AI). Of late AI-enabled methods such as machine learning (ML) and deep learning paved way for solving many real-world problems. As it is a learning-based approach, accumulated training samples help in improving thequality of training and thus leveraging malware detection accuracy. Existing deep learning methods are focusing on learning-based malware detection systems. However, there is need for improving the state of the art through ensemble approach. Towards this end, in this paper we proposed a framework known as Deep Ensemble Framework (DEF) for automatic malware detection. The framework obtains features from training samples. From given malware instance a grayscale image is generated. There is another process to extract the opcode sequences. Convolutional Neural Network (CNN) and Long Short Term Memory (LSTM) techniques are used to obtain grayscale image and opcode sequence respectively. Afterwards, a stacking ensemble is employed in order to achieve efficient malware detection and classification. Malware samples collected fromthe Internet sources and Microsoft are used for theempirical study. An algorithm known as Ensemble Learning for Automatic Malware Detection (EL-AML) is proposed to realize our framework. Another algorithm named Pre-Process is proposed to assist the EL-AML algorithm for obtaining intermediate features required by CNN and LSTM.Empirical study reveals that our framework outperforms many existing methods in terms of speed-up and accuracy.
High Performance NMF based Intrusion Detection System for Big Data IoT TrafficIJCNCJournal
With the emergence of smart devices and the Internet of Things (IoT), millions of users connected to the network produce massive network traffic datasets. These vast datasets of network traffic, Big Data are challenging to store, deal with and analyse using a single computer. In this paper we developed parallel implementation using a High Performance Computer (HPC) for the Non-Negative Matrix Factorization technique as an engine for an Intrusion Detection System (HPC-NMF-IDS). The large IoT traffic datasets of order of millions samples are distributed evenly on all the computing cores for both storage and speedup purpose. The distribution of computing tasks involved in the Matrix Factorization takes into account the reduction of the communication cost between the computing cores. The experiments we conducted on the proposed HPC-IDS-NMF give better results than the traditional ML-based intrusion detection systems. We could train the HPC model with datasets of one million samples in only 31 seconds instead of the 40 minutes using one processor), that is a speed up of 87 times. Moreover, we have got an excellent detection accuracy rate of 98% for KDD dataset.
IoT Guardian: A Novel Feature Discovery and Cooperative Game Theory Empowered...IJCNCJournal
Cyber intrusion attacks increasingly target the Internet of Things (IoT) ecosystem, exploiting vulnerable devices and networks. Malicious activities must be identified early to minimize damage and mitigate threats. Using actual benign and attack traffic from the CICIoT2023 dataset, this WORK aims to evaluate and benchmark machine-learning techniques for IoT intrusion detection. There are four main phases to the system. First, the CICIoT2023 dataset is refined to remove irrelevant features and clean up missing and duplicate data. The second phase employs statistical models and artificial intelligence to discover novel features. The most significant features are then selected in the third phase based on cooperative game theory. Using the original CICIoT2023 dataset and a dataset containing only novel features, we train and evaluate a variety of machine learning classifiers. On the original dataset, Random Forest achieved the highest accuracy of 99%. Still, with novel features, Random Forest's performance dropped only slightly (96%) while other models achieved significantly lower accuracy. As a whole, the work contributes substantial contributions to tailored feature engineering, feature selection, and rigorous benchmarking of IoT intrusion detection techniques. IoT networks and devices face continuously evolving threats, making it necessary to develop robust intrusion detection systems.
Enhancing Traffic Routing Inside a Network through IoT Technology & Network C...IJCNCJournal
IoT networking uses real items as stationary or mobile nodes. Mobile nodes complicate networking. Internet of Things (IoT) networks have a lot of control overhead messages because devices are mobile. These signals are generated by the constant flow of control data as such device identity, geographical positioning, node mobility, device configuration, and others. Network clustering is a popular overhead communication management method. Many cluster-based routing methods have been developed to address system restrictions. Node clustering based on the Internet of Things (IoT) protocol, may be used to cluster all network nodes according to predefined criteria. Each cluster will have a Smart Designated Node. SDN cluster management is efficient. Many intelligent nodes remain in the network. The network design spreads these signals. This paper presents an intelligent and responsive routing approach for clustered nodes in IoT networks. An existing method builds a new sub-area clustered topology. The Nodes Clustering Based on the Internet of Things (NCIoT) method improves message transmission between any two nodes. This will facilitate the secure and reliable interchange of healthcare data between professionals and patients. NCIoT is a system that organizes nodes in the Internet of Things (IoT) by grouping them together based on their proximity. It also picks SDN routes for these nodes. This approach involves selecting one option from a range of choices and preparing for likely outcomes problem addressing limitations on activities is a primary focus during the review process. Predictive inquiry employs the process of analyzing data to forecast and anticipate future events. This document provides an explanation of compact units. The Predictive Inquiry Small Packets (PISP) improved its backup system and partnered with SDN to establish a routing information table for each intelligent node, resulting in higher routing performance. Both principal and secondary roads are available for use. The simulation findings indicate that NCIoT algorithms outperform CBR protocols. Enhancements lead to a substantial 78% boost in network performance. In addition, the end-to-end latency dropped by 12.5%. The PISP methodology produces 5.9% more inquiry packets compared to alternative approaches. The algorithms are constructed and evaluated against academic ones.
IoT Guardian: A Novel Feature Discovery and Cooperative Game Theory Empowered...IJCNCJournal
Cyber intrusion attacks increasingly target the Internet of Things (IoT) ecosystem, exploiting vulnerable devices and networks. Malicious activities must be identified early to minimize damage and mitigate threats. Using actual benign and attack traffic from the CICIoT2023 dataset, this WORK aims to evaluate and benchmark machine-learning techniques for IoT intrusion detection. There are four main phases to the system. First, the CICIoT2023 dataset is refined to remove irrelevant features and clean up missing and duplicate data. The second phase employs statistical models and artificial intelligence to discover novel features. The most significant features are then selected in the third phase based on cooperative game theory. Using the original CICIoT2023 dataset and a dataset containing only novel features, we train and evaluate a variety of machine learning classifiers. On the original dataset, Random Forest achieved the highest accuracy of 99%. Still, with novel features, Random Forest's performance dropped only slightly (96%) while other models achieved significantly lower accuracy. As a whole, the work contributes substantial contributions to tailored feature engineering, feature selection, and rigorous benchmarking of IoT intrusion detection techniques. IoT networks and devices face continuously evolving threats, making it necessary to develop robust intrusion detection systems.
** Connect, Collaborate, And Innovate: IJCNC - Where Networking Futures Take ...IJCNCJournal
The International Journal of Computer Networks & Communications (IJCNC) is a bi monthly open access peer-reviewed journal that publishes articles which contribute new results in all areas of Computer Networks & Communications. The journal focuses on all technical and practical aspects of Computer Networks & data Communications. The goal of this journal is to bring together researchers and practitioners from academia and industry to focus on advanced networking concepts and establishing new collaborations in these areas.
Enhancing Traffic Routing Inside a Network through IoT Technology & Network C...IJCNCJournal
IoT networking uses real items as stationary or mobile nodes. Mobile nodes complicate networking. Internet of Things (IoT) networks have a lot of control overhead messages because devices are mobile. These signals are generated by the constant flow of control data as such device identity, geographical positioning, node mobility, device configuration, and others. Network clustering is a popular overhead communication management method. Many cluster-based routing methods have been developed to address system restrictions. Node clustering based on the Internet of Things (IoT) protocol, may be used to cluster all network nodes according to predefined criteria. Each cluster will have a Smart Designated Node. SDN cluster management is efficient. Many intelligent nodes remain in the network. The network design spreads these signals. This paper presents an intelligent and responsive routing approach for clustered nodes in IoT networks. An existing method builds a new sub-area clustered topology. The Nodes Clustering Based on the Internet of Things (NCIoT) method improves message transmission between any two nodes. This will facilitate the secure and reliable interchange of healthcare data between professionals and patients. NCIoT is a system that organizes nodes in the Internet of Things (IoT) by grouping them together based on their proximity. It also picks SDN routes for these nodes. This approach involves selecting one option from a range of choices and preparing for likely outcomes problem addressing limitations on activities is a primary focus during the review process. Predictive inquiry employs the process of analyzing data to forecast and anticipate future events. This document provides an explanation of compact units. The Predictive Inquiry Small Packets (PISP) improved its backup system and partnered with SDN to establish a routing information table for each intelligent node, resulting in higher routing performance. Both principal and secondary roads are available for use. The simulation findings indicate that NCIoT algorithms outperform CBR protocols. Enhancements lead to a substantial 78% boost in network performance. In addition, the end-to-end latency dropped by 12.5%. The PISP methodology produces 5.9% more inquiry packets compared to alternative approaches. The algorithms are constructed and evaluated against academic ones.
How to Make a Field invisible in Odoo 17Celine George
It is possible to hide or invisible some fields in odoo. Commonly using “invisible” attribute in the field definition to invisible the fields. This slide will show how to make a field invisible in odoo 17.
Introduction to AI for Nonprofits with Tapp NetworkTechSoup
Dive into the world of AI! Experts Jon Hill and Tareq Monaur will guide you through AI's role in enhancing nonprofit websites and basic marketing strategies, making it easy to understand and apply.
Unit 8 - Information and Communication Technology (Paper I).pdfThiyagu K
This slides describes the basic concepts of ICT, basics of Email, Emerging Technology and Digital Initiatives in Education. This presentations aligns with the UGC Paper I syllabus.
A Strategic Approach: GenAI in EducationPeter Windle
Artificial Intelligence (AI) technologies such as Generative AI, Image Generators and Large Language Models have had a dramatic impact on teaching, learning and assessment over the past 18 months. The most immediate threat AI posed was to Academic Integrity with Higher Education Institutes (HEIs) focusing their efforts on combating the use of GenAI in assessment. Guidelines were developed for staff and students, policies put in place too. Innovative educators have forged paths in the use of Generative AI for teaching, learning and assessments leading to pockets of transformation springing up across HEIs, often with little or no top-down guidance, support or direction.
This Gasta posits a strategic approach to integrating AI into HEIs to prepare staff, students and the curriculum for an evolving world and workplace. We will highlight the advantages of working with these technologies beyond the realm of teaching, learning and assessment by considering prompt engineering skills, industry impact, curriculum changes, and the need for staff upskilling. In contrast, not engaging strategically with Generative AI poses risks, including falling behind peers, missed opportunities and failing to ensure our graduates remain employable. The rapid evolution of AI technologies necessitates a proactive and strategic approach if we are to remain relevant.
June 3, 2024 Anti-Semitism Letter Sent to MIT President Kornbluth and MIT Cor...Levi Shapiro
Letter from the Congress of the United States regarding Anti-Semitism sent June 3rd to MIT President Sally Kornbluth, MIT Corp Chair, Mark Gorenberg
Dear Dr. Kornbluth and Mr. Gorenberg,
The US House of Representatives is deeply concerned by ongoing and pervasive acts of antisemitic
harassment and intimidation at the Massachusetts Institute of Technology (MIT). Failing to act decisively to ensure a safe learning environment for all students would be a grave dereliction of your responsibilities as President of MIT and Chair of the MIT Corporation.
This Congress will not stand idly by and allow an environment hostile to Jewish students to persist. The House believes that your institution is in violation of Title VI of the Civil Rights Act, and the inability or
unwillingness to rectify this violation through action requires accountability.
Postsecondary education is a unique opportunity for students to learn and have their ideas and beliefs challenged. However, universities receiving hundreds of millions of federal funds annually have denied
students that opportunity and have been hijacked to become venues for the promotion of terrorism, antisemitic harassment and intimidation, unlawful encampments, and in some cases, assaults and riots.
The House of Representatives will not countenance the use of federal funds to indoctrinate students into hateful, antisemitic, anti-American supporters of terrorism. Investigations into campus antisemitism by the Committee on Education and the Workforce and the Committee on Ways and Means have been expanded into a Congress-wide probe across all relevant jurisdictions to address this national crisis. The undersigned Committees will conduct oversight into the use of federal funds at MIT and its learning environment under authorities granted to each Committee.
• The Committee on Education and the Workforce has been investigating your institution since December 7, 2023. The Committee has broad jurisdiction over postsecondary education, including its compliance with Title VI of the Civil Rights Act, campus safety concerns over disruptions to the learning environment, and the awarding of federal student aid under the Higher Education Act.
• The Committee on Oversight and Accountability is investigating the sources of funding and other support flowing to groups espousing pro-Hamas propaganda and engaged in antisemitic harassment and intimidation of students. The Committee on Oversight and Accountability is the principal oversight committee of the US House of Representatives and has broad authority to investigate “any matter” at “any time” under House Rule X.
• The Committee on Ways and Means has been investigating several universities since November 15, 2023, when the Committee held a hearing entitled From Ivory Towers to Dark Corners: Investigating the Nexus Between Antisemitism, Tax-Exempt Universities, and Terror Financing. The Committee followed the hearing with letters to those institutions on January 10, 202
Macroeconomics- Movie Location
This will be used as part of your Personal Professional Portfolio once graded.
Objective:
Prepare a presentation or a paper using research, basic comparative analysis, data organization and application of economic information. You will make an informed assessment of an economic climate outside of the United States to accomplish an entertainment industry objective.
Honest Reviews of Tim Han LMA Course Program.pptxtimhan337
Personal development courses are widely available today, with each one promising life-changing outcomes. Tim Han’s Life Mastery Achievers (LMA) Course has drawn a lot of interest. In addition to offering my frank assessment of Success Insider’s LMA Course, this piece examines the course’s effects via a variety of Tim Han LMA course reviews and Success Insider comments.
Instructions for Submissions thorugh G- Classroom.pptxJheel Barad
This presentation provides a briefing on how to upload submissions and documents in Google Classroom. It was prepared as part of an orientation for new Sainik School in-service teacher trainees. As a training officer, my goal is to ensure that you are comfortable and proficient with this essential tool for managing assignments and fostering student engagement.
Synthetic Fiber Construction in lab .pptxPavel ( NSTU)
Synthetic fiber production is a fascinating and complex field that blends chemistry, engineering, and environmental science. By understanding these aspects, students can gain a comprehensive view of synthetic fiber production, its impact on society and the environment, and the potential for future innovations. Synthetic fibers play a crucial role in modern society, impacting various aspects of daily life, industry, and the environment. ynthetic fibers are integral to modern life, offering a range of benefits from cost-effectiveness and versatility to innovative applications and performance characteristics. While they pose environmental challenges, ongoing research and development aim to create more sustainable and eco-friendly alternatives. Understanding the importance of synthetic fibers helps in appreciating their role in the economy, industry, and daily life, while also emphasizing the need for sustainable practices and innovation.
Francesca Gottschalk - How can education support child empowerment.pptxEduSkills OECD
Francesca Gottschalk from the OECD’s Centre for Educational Research and Innovation presents at the Ask an Expert Webinar: How can education support child empowerment?
Overview on Edible Vaccine: Pros & Cons with Mechanism
A STRUCTURED DEEP NEURAL NETWORK FOR DATA-DRIVEN LOCALIZATION IN HIGH FREQUENCY WIRELESS NETWORKS
1. International Journal of Computer Networks & Communications (IJCNC) Vol.9, No.3, May 2017
DOI: 10.5121/ijcnc.2017.9302 21
A STRUCTURED DEEP NEURAL NETWORK FOR
DATA-DRIVEN LOCALIZATION IN HIGH
FREQUENCY WIRELESS NETWORKS
Marcus Z. Comiter1
, Michael B. Crouse1
and H. T. Kung1
1
John A. Paulson School of Engineering and Applied Science
Harvard University
Cambridge, MA 02138
marcuscomiter@g.harvard.edu
mcrouse@g.harvard.edu
kung@harvard.edu
ABSTRACT
Next-generation wireless networks such as 5G and 802.11ad networks will use millimeter waves operating
at 28GHz, 38GHz, or higher frequencies to deliver unprecedentedly high data rates, e.g., 10 gigabits per
second. However, millimeter waves must be used directionally with narrow beams in order to overcome the
large attenuation due to their higher frequency. To achieve high data rates in a mobile setting,
communicating nodes need to align their beams dynamically, quickly, and in high resolution. We propose a
data-driven, deep neural network (DNN) approach to provide robust localization for beam alignment,
using a lower frequency spectrum (e.g., 2.4 GHz). The proposed DNN-based localization methods use the
angle of arrival derived from phase differences in the signal received at multiple antenna arrays to infer the
location of a mobile node. Our methods differ from others that use DNNs as a black box in that the
structure of our neural network model is tailored to address difficulties associated with the domain, such as
collinearity of the mobile node with antenna arrays, fading and multipath. We show that training our
models requires a small number of sample locations, such as 30 or fewer, making the proposed methods
practical. Our specific contributions are: (1) a structured DNN approach where the neural network
topology reflects the placement of antenna arrays, (2) a simulation platform for generating training and
evaluation data sets under multiple noise models, and (3) demonstration that our structured DNN approach
improves localization under noise by up to 25% over traditional off-the-shelf DNNs, and can achieve sub-
meter accuracy in a real-world experiment.
KEYWORDS
Millimeter wave, 5G, 802.11ad, Localization, Mobile networks, Machine learning, Deep Neural Networks
1. INTRODUCTION
The next generation of wireless networks will utilize high-frequency millimeter waves
(mmWaves). A consequence of this is the need for accurate narrow-beam alignment between
receiving and transmitting nodes [1]. One such alignment method is to localize mobile nodes
such that base station antennas can align their narrow beams (e.g., under 30 degrees) towards
mobile nodes, and vice versa, when mobile nodes also possess directional antennas. This need
for localization is reflected in the emerging 5G standard, which will use mmWave spectrum and
explicitly calls for sub-meter localization accuracy for indoor deployments [2]. We propose new,
data-driven localization methods using lower frequency spectrum such as 2.4GHz or 916MHz, as
2. International Journal of Computer Networks & Communications (IJCNC) Vol.9, No.3, May 2017
22
in the experiments in this paper, and their omni- or quasi-directional functionality to provide a
low-overhead, accurate localization to be used for aligning higher frequency end points.
While there exist many methods for localization, there are a number of domain-specific
challenges related to the goal of providing accurate beam alignment for mmWave wireless
networks. First, the localization must be highly accurate if it is to be used for the purposes of
aligning narrow beams without significant link loss. Second, as the beam alignment procedure
must be executed in settings in which the location of the mobile node is rapidly changing, the
localization procedure must be sufficiently fast. Third, the localization procedure must be well
suited to the challenges posed by the environments and sensing mechanisms in which the
networks will operate. For example, indoor environments are highly susceptible to dynamic
multipath effects due to the propagation of electromagnetic waves and their interaction with the
high density and diversity of objects and materials within the environments.
To address these challenges, we propose a fully data-driven neural network approach to localizing
mobile nodes for the purposes of aligning antennas of communicating devices accurately and in
low latency. Our proposed approach is able to localize mobile nodes using signal patterns learned
specifically from the data associated with the environment, making our method particularly well
suited to complicated multipath environments that are difficult to capture analytically using
traditional channel models. Importantly, as our method requires only data to train the localization
model, it is not necessary to have any explicit rules or heuristics to function in these complicated
environments. For these reasons, our data-driven method is applicable to the highly
heterogeneous individual environments in which mmWave systems will be deployed. For a given
environment, only a small number of location samples is required (e.g., fewer than 30) for
learning the localization neural network.Further, we show that a relatively small deep neural
network (DNN) model (e.g., with four layers) is sufficient for accurate localization, and can infer
locations in under a millisecond via a feed-forward operation.
We propose novel neural network techniques specifically designed for the high frequency
wireless domain and associated antenna alignment application requirements while simultaneously
minimizing the amount of real-world training samples that must be collected. Note that in order
for such a data-driven approach to be practically applicable, it must be able to operate without
requiring overly onerous amounts of data collection for new environments. Further, a fully data-
driven approach must automatically address multipath effects, occlusions, and other
characteristics of the environment that influence the behavior of electromagnetic waves. Our
proposed machine learning algorithms are able to learn these complicated effects automatically.
Our main contributions are as follows:
1. We introduce a structured neural network model to address collinearity regions between
base stationswith antenna arrays that capture phase differences in received signal from a
mobile node to be localized.
2. To minimize the data collection effort, we introduce a simulator for generating training
data for the localization neural networkunder various scenarios including differing
number of base stations, number of sample points, and under multiple noise models.
3. We demonstrate with both simulated data and real-world experiments that our structured
DNN approach improves localization under all noise models considered by up to 25%
over traditional off-the-shelf DNNs. Specifically, we achieve sub-meter in both
simulated indoor and real-world experiments.
3. International Journal of Computer Networks & Communications (IJCNC) Vol.9, No.3, May 2017
23
2. RELATED WORK
Our work focuses on localizing mobile nodesusing multiple base stations each equipped
withantenna arrays in a lower frequency, such as 2.4 GHz. Earlier attempts at localizing nodes
with wireless network connections include using signals such as Received Signal Strength
Indicator (RSSI) (see, e.g., [16]). RSSI is applicable to nodes using a number of different
networking protocols, including Wi-Fi [18] and Bluetooth [17]. However, while RSSI can be
used to calculate a coarse estimate of distance between a receiving and transmitting node, in
practice the granularity of the localization is not sufficient to meet a goal of sub-meter fidelity
[16].
In contrast to directly using RSSI provided by the radio signals, our method uses additional
information in signals to localize nodes beyond its signal strength. There are a number of other
methods that approach the problem similarly. Adib et al. [4,5] present methods for tracking users
through walls using frequency modulated carrier waves (FMCW), or a chirp signal, and multiple
users moving within an indoor environment can be localized and tracked. FMCW requires
specialized hardware to generate the “chirp” signal that enables the time of flight measurements
to be calculated accurately and then used in determining the location. This method relies on the
users to be moving in order to separate their reflections from the static background, is limited to
just a few users, and requires a windowed set of measurements for accurate background
subtraction.
A motivation for our work comes more directly from the 802.11ad Amendment III, which
addresses the next generation of indoor wireless networking, focusing on connecting and using
high-frequency networks [6]. The amendment proposes to perform beamforming and alignment
through a procedure called Sector Level Sweep (SLS), an in-band search to locate the correct
beam sectors to utilize for inter-device communication. SLS scans over the available antenna
sectors, and is then refined via the Beam Refinement Process (BRP). To make this process more
efficient, Blind Beam Steering (BBS) [7] has been proposed to reduce the amount of scanning by
utilizing a lower frequency, 2.4 GHz, and computing an angle-of-arrival using MUSIC.
Our approach is similarly motivated by the desire to reduce the amount of scanning by utilizing a
lower frequency mode for localization purposes. However, while BBS may need to revert to an
SLS when large multipath effects are detected, our data-driven methods instead allow our model
to learn multipath effects from both observed and synthetic data, and work effectively and
automatically in these regions. Further, by calculating a position rather than just the angle-of-
arrival, our method can quickly find one of the alternative beams associated with the location of
the mobile node for use during intermittent blockages, doing so without the cost of beam
sweeping. In addition, the localization information computed by our method can be useful not
only for beam alignment purposes, but also for antenna selection and future medium access
control algorithms.
ArrayTrack is a system that utilizes a set of base stations equipped with antenna arrays throughout
an environment in order to localize mobile nodes within the environment using angle-of-arrival
measurements [8]. This work utilizes as many as eight antennas on each base station in order to
limit the error of the MUSIC algorithm when calculating the angle-of-arrival, specifically
attempting to improve MUSIC's robustness to the multipath effects plaguing indoor wireless
environments.
In the area of using data-driven methods for localization, deep neural networks for performing
localization have been previously evaluated under multiple indoor scenarios [9,10,11].
Our method improves upon the localization error presented in the aforementioned work by
4. International Journal of Computer Networks & Communications (IJCNC) Vol.
modifying the neural network structure (topology) to suit the application domain.
data-driven method is robust to multipath effects by learning models
measurements that contain these effects. As such, we show our method can provide better
accuracy using fewer antennas and base stations in our real
sub-meter fidelity.
3. CHALLENGES FOR LOCALIZATION
We illustrate the operating scenario for the overarching localization problem for indoor or
outdoor networks operating with dual modes in Figure 1. We consider an environment in which a
user is operating a mobile node that may change position over time. The
wirelessly by a number of base stations that are dispersed throughout the environment. We
assume that each base station contains multiple antennas, with one array operating at out
low frequencies (such as 2.4 GHz in Wi
another array operating at higher mmWave frequencies (such as 28, 38, or 60 GHz).
that for the rest of the paper, we will use 2.4 GHz as an example of lower frequencies used for the
localization purpose.) In this scenario, it is desirable for the mobile node to communicate with a
given base station over the high frequency mmWave channel in order to achieve higher data rates
as compared with lower frequency links. However, due to the directional natu
antennas and systems, it is necessary for the endpoints to know in which direction, or sector, to
steer the beam. This can be accomplished directly with knowledge of the location of the mobile
node within an environment.
There are a number of challenges associated with localizing mobile nodes within a realistic
environment in order to meet specific localization requirements for high data
communications. These challenges stem from both the system itself, such as collinearity
problems induced by the placement of base stations, as well as the properties of electromagnetic
waves within the environment, such as multipath and fading effects. We now discuss these
challenges in detail.
Figure 1: Localization scenario for mmWave n
high-speed data transmission and 2.4 GHz frequency band for localization. We also illustrate the manner in
which angle
International Journal of Computer Networks & Communications (IJCNC) Vol.9, No.3, May 2017
modifying the neural network structure (topology) to suit the application domain. O
driven method is robust to multipath effects by learning models directly from real
measurements that contain these effects. As such, we show our method can provide better
accuracy using fewer antennas and base stations in our real-world experiments demonstrating
OCALIZATION
illustrate the operating scenario for the overarching localization problem for indoor or
outdoor networks operating with dual modes in Figure 1. We consider an environment in which a
user is operating a mobile node that may change position over time. The mobile node is served
wirelessly by a number of base stations that are dispersed throughout the environment. We
assume that each base station contains multiple antennas, with one array operating at out
low frequencies (such as 2.4 GHz in Wi-Fi and 916 MHz in our real-world experiments) and
another array operating at higher mmWave frequencies (such as 28, 38, or 60 GHz).
that for the rest of the paper, we will use 2.4 GHz as an example of lower frequencies used for the
In this scenario, it is desirable for the mobile node to communicate with a
given base station over the high frequency mmWave channel in order to achieve higher data rates
as compared with lower frequency links. However, due to the directional nature of mmWave
antennas and systems, it is necessary for the endpoints to know in which direction, or sector, to
steer the beam. This can be accomplished directly with knowledge of the location of the mobile
f challenges associated with localizing mobile nodes within a realistic
environment in order to meet specific localization requirements for high data-rate mmWave
communications. These challenges stem from both the system itself, such as collinearity
ems induced by the placement of base stations, as well as the properties of electromagnetic
waves within the environment, such as multipath and fading effects. We now discuss these
for mmWave networking. In this illustration, we use 60 GHz mmWave for
speed data transmission and 2.4 GHz frequency band for localization. We also illustrate the manner in
which angle-of-arrival is measured, denoted by θ.
9, No.3, May 2017
24
Our proposed
directly from real
measurements that contain these effects. As such, we show our method can provide better
world experiments demonstrating
illustrate the operating scenario for the overarching localization problem for indoor or
outdoor networks operating with dual modes in Figure 1. We consider an environment in which a
mobile node is served
wirelessly by a number of base stations that are dispersed throughout the environment. We
assume that each base station contains multiple antennas, with one array operating at out-of-band
world experiments) and
another array operating at higher mmWave frequencies (such as 28, 38, or 60 GHz). (Please note
that for the rest of the paper, we will use 2.4 GHz as an example of lower frequencies used for the
In this scenario, it is desirable for the mobile node to communicate with a
given base station over the high frequency mmWave channel in order to achieve higher data rates
re of mmWave
antennas and systems, it is necessary for the endpoints to know in which direction, or sector, to
steer the beam. This can be accomplished directly with knowledge of the location of the mobile
f challenges associated with localizing mobile nodes within a realistic
rate mmWave
communications. These challenges stem from both the system itself, such as collinearity
ems induced by the placement of base stations, as well as the properties of electromagnetic
waves within the environment, such as multipath and fading effects. We now discuss these
etworking. In this illustration, we use 60 GHz mmWave for
speed data transmission and 2.4 GHz frequency band for localization. We also illustrate the manner in
5. International Journal of Computer Networks & Communications (IJCNC) Vol.
3.1. Collinearity
Within the context of localizing a mobile node, we define a
base stations as the region along a straight line connecting the two base stations, as shown in
Figure 2. From the perspective of localizing a mobile node, this
as the location of any node located along this line cannot be uniquely determined from the angles
of-arrival between the node and the two base stations or the phase offsets at the two base stations.
Figure 2: An illustration of the collinearity issue, which the
localization neural network of this paper addresses
In the context of data-driven machine learning, the fact that there exists one input (either angles
of-arrival or phase offsets) that maps to multiple outputs (multiple locations within the collinear
region) introduces a one-to-many relationship. This mak
location of a mobile node correctly within the collinear region. Within the collinear region
between two base stations, the model
large amount of error).
More specifically, the model tends to predict any node within the collinear region as being at the
midpoint of the collinear region. This stems from the fact that minimizing the loss for a one
many mapping is accomplished by predicting the output
points whose inputs are the same, which can be shown to be the maximum likelihood estimate.
While having greater than two base stations can resolve these regions analytically, data
conventional machine learning models without our proposed modifications to the neural
network’s structure that reflect the placement of base station nodes have trouble resolving these
regions under noise, or require much larger amounts of training data even when noise is not
present.
3.2. Multipath
In realistic deployment scenarios, the presence of walls, reflective surfaces, objects, and other
physical features that interact with the propagation of electromagnetic waves can result in
International Journal of Computer Networks & Communications (IJCNC) Vol.9, No.3, May 2017
Within the context of localizing a mobile node, we define a collinear region between a pair of
base stations as the region along a straight line connecting the two base stations, as shown in
Figure 2. From the perspective of localizing a mobile node, this collinear region is of special note,
as the location of any node located along this line cannot be uniquely determined from the angles
arrival between the node and the two base stations or the phase offsets at the two base stations.
stration of the collinearity issue, which the Structured Multi-Layer Perceptron (SMLP)
localization neural network of this paper addresses.
driven machine learning, the fact that there exists one input (either angles
arrival or phase offsets) that maps to multiple outputs (multiple locations within the collinear
many relationship. This makes it impossible to accurately infer the
location of a mobile node correctly within the collinear region. Within the collinear region
between two base stations, the model would therefore predict the location incorrectly (i.e., with a
More specifically, the model tends to predict any node within the collinear region as being at the
midpoint of the collinear region. This stems from the fact that minimizing the loss for a one
many mapping is accomplished by predicting the output as the average of the outputs of all data
points whose inputs are the same, which can be shown to be the maximum likelihood estimate.
While having greater than two base stations can resolve these regions analytically, data
rning models without our proposed modifications to the neural
network’s structure that reflect the placement of base station nodes have trouble resolving these
regions under noise, or require much larger amounts of training data even when noise is not
In realistic deployment scenarios, the presence of walls, reflective surfaces, objects, and other
physical features that interact with the propagation of electromagnetic waves can result in
9, No.3, May 2017
25
between a pair of
base stations as the region along a straight line connecting the two base stations, as shown in
collinear region is of special note,
as the location of any node located along this line cannot be uniquely determined from the angles-
arrival between the node and the two base stations or the phase offsets at the two base stations.
Layer Perceptron (SMLP)
driven machine learning, the fact that there exists one input (either angles-
arrival or phase offsets) that maps to multiple outputs (multiple locations within the collinear
es it impossible to accurately infer the
location of a mobile node correctly within the collinear region. Within the collinear region
the location incorrectly (i.e., with a
More specifically, the model tends to predict any node within the collinear region as being at the
midpoint of the collinear region. This stems from the fact that minimizing the loss for a one-to-
as the average of the outputs of all data
points whose inputs are the same, which can be shown to be the maximum likelihood estimate.
While having greater than two base stations can resolve these regions analytically, data-driven,
rning models without our proposed modifications to the neural
network’s structure that reflect the placement of base station nodes have trouble resolving these
regions under noise, or require much larger amounts of training data even when noise is not
In realistic deployment scenarios, the presence of walls, reflective surfaces, objects, and other
physical features that interact with the propagation of electromagnetic waves can result in
6. International Journal of Computer Networks & Communications (IJCNC) Vol.9, No.3, May 2017
26
multipath effects. Within the context of localization, multipath interference can often introduce
large variations when estimating the angle-of-arrival, as it will alter the phase or time of flight
measured at the base stations.
Without accounting for multipath, which can be difficult when building explicit models of the
room or even using statistical models such as MUSIC, the derived location of the mobile node
can be very inaccurate. While advancements in these statistical methods have been made to
better address this, problems still exist in the presence of large multipath effects [7,8]. For indoor
deployments of mmWave networks, it is imperative that the localization methods are able to
operate under multipath effectsresulting from reflections off of various types of materials. This is
primarily due to the great variety of objects, materials, and architectural layouts common to
deployment environments. In our proposed data-driven approach, rather than accounting for
these effects or attempting to model them explicitly, small amounts of data can be sampled from
the room and used to train a model that captures the key propagation characteristics of the given
scenario. As long as regions of the environment with large multipath effects are sampled, these
key characteristics can be learned by the model.
3.3. Latency and Interference
The procedure for selecting which beam sector to use, or beam to form, may be executed in
settings in which the location of the mobile node is rapidly changing. As a result, any localization
procedure must be sufficiently fast to meet these conditions. When the alignment latency
between end points is too long, the system initialization overhead will be high, while intermittent
latency (e.g., MAC ARQ) resulting from packet loss due to misaligned beams can potentially
disrupt higher layer protocols, causing loss of throughput from, for example, TCP timeouts [12].
Further, as many new applications have stringent latency budgets, such as Virtual Reality (VR)
and Augmented Reality (AR), it is important that deployments requiring localization are
cognizant of the latency cost of the localization method used.
Some of the approaches to beamforming are particularly not well suited to the stringent latency
budgets of emerging applications. For example, blind in-band beam scanning in high
frequencies, in which the optimal antenna sector is found by scanning over many antenna sectors,
incurs delays that may make it incompatible with use in low-latency systems. In addition, this
scanning procedure can introduce interference, especially under multi-mobile node and multi-
base station scenarios, which further reduces its practicality.
In contrast to this, deep neural networks have recently seen significant improvement in
feedforward inference time and storage costs. For example, binarized neural networks, where all
the weights are 1-bit values, can be used even on memory and power constrained embedded
devices [13,14]. For such neural networks, the computation time for networks with fewer than
ten layers and small input sizes can be performed extremely efficiently and with latency under a
millisecond, as discussed in Section 5. As such, our proposed data-driven methods are cognizant
of the latency requirements of emerging applications.
Further, as our methods localize the node using a separate, lower frequency band rather than the
mmWave bands used for high speed data transmission, our method does not suffer from the
interference introduced by the beam scanning procedure. As such, our method can function
without interfering with other mobile nodes and base stations operating in the same environment,
as is expected to be the case in many mmWave deployments.
7. International Journal of Computer Networks & Communications (IJCNC) Vol.9, No.3, May 2017
27
4. DATA-DRIVEN LOCALIZATION METHODS
We now describe the specifics of our data-driven DNN-based approach for predicting the location
of a mobile node given measurement data at the base stations. We name our method the
Structured Multi-Layer Perceptron (SMLP). The SMLP is a structured learner and inference
engine consisting of multiple deep neural networks (which we define as neural networks with
more than one hidden layer). The SMLP, as we will show, improves the localization accuracy
and robustness for localization within collinear regions beyond the capabilities of off-the-shelf
DNNs.
4.1. The Structured Multilayer Perceptron Model
The Structured Multilayer Perceptron model (SMLP) is a type of DNN with domain specific
neural network modifications. At a high level, SMLPs are a combination of multiple DNNs
structured in a way such that the model can automatically address the challenge of localizing
mobile nodes within regions of collinearity. Without these modifications, conventional neural
networks such as a Multi-layer Perceptron (MLP) would find it difficult to accurately localize
mobile nodes in collinear regions without large amounts of training data.
We now detail the construction of the SMLP, which is illustrated in Figure 3. The SMLP is
constructed of two components, a Lower Pair Network (LPN) and an Upper Connection
Network (UCN), which both may contain a number of MLPs. Each MLP, both within the LPN
and the UCN, is formulated as a regression problem. Specifically, the output of each MLP is an
unrestricted physical location in the real space, which consists of all possible locations. This is in
contrast to formulating the training as a classification problem, which would instead have as
output a location chosen from a predetermined and finite set of locations, or bins.
The LPN consists of a combination of submodels, each of which are fully connected MLPs. Each
neural network in the LPN is trained on the data from pairs of base stations with collinear regions.
Importantly, each neural network in the LPN is trained independently of one another. Once
trained, each neural network in the LPN has the functionality to be able to take as input the angle-
of-arrivals to each of the base stations, and output a localization of the mobile node.
Note however, that due to the way each network is trained, these individual networks will not be
able to accurately localize nodes within the collinear region. This motivates the introduction of
the UCN, which takes the output from all of the LPN networks and uses it to resolve the collinear
problems. Specifically, we remove the final output layer of each network in the LPN, and
concatenate the output together. This is then used as the input to the UCN, which outputs the
final localization of the mobile node. By having a global view from multiple base station pair
combinations by proxy from the concatenated LPN output, the UCN is able to rectify the
collinearity problem, often with significantly less training examples.
4.2. Training SMLPs
We now discuss the training procedure for SMLPs for the wireless localization scenario. SMLPs
are trained offline on either collected or synthetic training data. On a 3.4 GHz CPU with 16GB of
RAM, the models used in all results in the paper can be trained in less than one minute. As
discussed previously, all training is formulated as a regression problem rather than a classification
problem, such that the final output of the SMLP is a location within the real space rather than a
probability distribution over predetermined locations.
All portions of the SMLP are trained using conventional neural network training methods, which
8. International Journal of Computer Networks & Communications (IJCNC) Vol.
we now briefly describe. To train a fully connected neural network with
x2, … xn-1, xn] neurons per hidden layer, initial weights are assigned to each neuron in each layer.
Paired (input, output) training samples are then
Each input, either individually or in a batch, is fed through the neural network, producing a final
output. This final output is then compared to the ground truth output of the training sample via a
loss function. For our models in this paper, we use an
function calculates the square-root of the squared difference between the output of the neural
network and the ground truth output. This error is then propagated
network using backpropagation, a process through which the weights of each of the neurons is
adjusted conditional on the amount of error. This process is repeated until convergence or until a
user-defined number of iterations is reac
We now discuss the training process for the Lower Pair Network (LPN), and afterwards describe
the training process for the Upper Connection Network (UCN). The training samples for each
neural network in the LPN are angles
of base stations. The output used in training the model is the location of the mobile node in the
real space. We use a rectified linear unit (ReLU) as the activation function, and an
function. Each neural network within the LPN is trained independently of one another
are no connections between each unit of the LPN, only in the UCN
We now describe the training process for the UCN. The input to the UCN training is derived
from the output of the LPN. Specifically, the last layer from each of the neural networks within
the LPN is removed. Following this, the output from the new final layer from each of the neural
networks in the LPN are concatenated to form a single long vector. This vector is us
training the UCN. The output used in training the UCN is the location of the mobile node in the
real space. We again use a ReLU as the activation function, and an
Note that each neural network within the LPN as well
one another. Specifically, any error in training the UCN is not back propagated to the LPN.
By combining the output of the multiple complementary LPN models at the UCN, the SMLP is
able to resolve the collinear regions that the lower layer submodels are unable to resolve by
themselves. As such, there is no need to backpropagate through the entire model, as the UCN
resolves the errors made in collinear regions by the LPN.
Figure 3: A diagram of our proposed Structured Multilayer Perceptron (SMLP) model, which separates
International Journal of Computer Networks & Communications (IJCNC) Vol.9, No.3, May 2017
we now briefly describe. To train a fully connected neural network with x hidden layers and [
] neurons per hidden layer, initial weights are assigned to each neuron in each layer.
Paired (input, output) training samples are then used to find the optimal weights for each neuron.
Each input, either individually or in a batch, is fed through the neural network, producing a final
output. This final output is then compared to the ground truth output of the training sample via a
function. For our models in this paper, we use an l2-norm loss function, such that the loss
root of the squared difference between the output of the neural
network and the ground truth output. This error is then propagated backwards through the
, a process through which the weights of each of the neurons is
adjusted conditional on the amount of error. This process is repeated until convergence or until a
defined number of iterations is reached.
We now discuss the training process for the Lower Pair Network (LPN), and afterwards describe
the training process for the Upper Connection Network (UCN). The training samples for each
neural network in the LPN are angles-of-arrivals (AOAs), as measured by the correspondin
output used in training the model is the location of the mobile node in the
real space. We use a rectified linear unit (ReLU) as the activation function, and an
network within the LPN is trained independently of one another
are no connections between each unit of the LPN, only in the UCN.
We now describe the training process for the UCN. The input to the UCN training is derived
PN. Specifically, the last layer from each of the neural networks within
the LPN is removed. Following this, the output from the new final layer from each of the neural
networks in the LPN are concatenated to form a single long vector. This vector is us
training the UCN. The output used in training the UCN is the location of the mobile node in the
real space. We again use a ReLU as the activation function, and an l2-norm loss function.
Note that each neural network within the LPN as well as the UCN are all trained independently of
one another. Specifically, any error in training the UCN is not back propagated to the LPN.
By combining the output of the multiple complementary LPN models at the UCN, the SMLP is
near regions that the lower layer submodels are unable to resolve by
themselves. As such, there is no need to backpropagate through the entire model, as the UCN
resolves the errors made in collinear regions by the LPN.
posed Structured Multilayer Perceptron (SMLP) model, which separates
9, No.3, May 2017
28
hidden layers and [x1,
] neurons per hidden layer, initial weights are assigned to each neuron in each layer.
used to find the optimal weights for each neuron.
Each input, either individually or in a batch, is fed through the neural network, producing a final
output. This final output is then compared to the ground truth output of the training sample via a
norm loss function, such that the loss
root of the squared difference between the output of the neural
backwards through the
, a process through which the weights of each of the neurons is
adjusted conditional on the amount of error. This process is repeated until convergence or until a
We now discuss the training process for the Lower Pair Network (LPN), and afterwards describe
the training process for the Upper Connection Network (UCN). The training samples for each
asured by the corresponding pairs
output used in training the model is the location of the mobile node in the
real space. We use a rectified linear unit (ReLU) as the activation function, and an l2-norm loss
network within the LPN is trained independently of one another, as there
We now describe the training process for the UCN. The input to the UCN training is derived
PN. Specifically, the last layer from each of the neural networks within
the LPN is removed. Following this, the output from the new final layer from each of the neural
networks in the LPN are concatenated to form a single long vector. This vector is used as input in
training the UCN. The output used in training the UCN is the location of the mobile node in the
norm loss function.
as the UCN are all trained independently of
one another. Specifically, any error in training the UCN is not back propagated to the LPN.
By combining the output of the multiple complementary LPN models at the UCN, the SMLP is
near regions that the lower layer submodels are unable to resolve by
themselves. As such, there is no need to backpropagate through the entire model, as the UCN
posed Structured Multilayer Perceptron (SMLP) model, which separates
9. International Journal of Computer Networks & Communications (IJCNC) Vol.
base stations pairs in the Lower Pair Network (LPN) and then combines them in the Upper Connection
Network (UCN). The SMLP infers the (x, y)
5. SIMULATION AND RESULTS
In order to validate our proposed methods, we have built a detailed and realistic simulator, and
use it to generate data with which our methods can be evaluated. We first describe the simulator
we have built for these purposes. We next present results us
different conditions, including complex noise models validated through real
For all results in this section, we use a virtual setup with three base stations, and evaluate on a
held-out test set of 2500 points canvassing the entire 50x50 virtual space corresponding to a 3m
by 3m grid. All results are reported in terms of Median Square Error (Median SE).
Before describing results, we briefly comment on the latency ramifications of SMLPs. Note that
the SMLP models in this section use 2 layers in the Lower Pair Network (LPN) with 500 and 50
neurons per layer, respectively, and 2 layers in the Upper Connection Network with 200 and 50
neurons per layer, respectively. Further, the number of multiply and a
to use the SMLP for inference can be found according to the following formula:
L - # of LPNs
dl- dimension of LPN input
n1k- # of neurons in LPN layer k
U - # of UCNs
du - dimension of UCN input
nuk- # of neurons in UCN layer k
Therefore, for the SMLP used in this paper, a total of
multiply and addition operations are required. A commodity accelerator capable of executing
4500 GFLOPs can execute these operations in 0.01 milliseconds, making our method applicable
in scenarios with small latency budgets.
5.1. Simulator
In order to evaluate our proposed SMLP Model, we have created a simulator for generating
training data under a number of different scenarios of interest, including equipment failure and
noise. We have released the simulator publicly as a community resource. The code for the
simulator can be found online and is under active development (currently available at
https://github.com/KevinHCChen/wireless
The simulator first constructs a configurable virt
simulator then places mobile nodes within the space, as well as base stations. Given this setup,
the simulator generates the data needed to train the SMLP models. Specifically, based on the
relative location of the base stations and the mobile nodes, the simulator generates either time of
flight or angle-of-arrival data to each of the base stations given a virtual mobile node and its
position, and pairs this with the location of the mobile nodes. This da
it can be used in training the SMLP models.
International Journal of Computer Networks & Communications (IJCNC) Vol.9, No.3, May 2017
base stations pairs in the Lower Pair Network (LPN) and then combines them in the Upper Connection
Network (UCN). The SMLP infers the (x, y)-coordinates of a mobile device.
ESULTS
In order to validate our proposed methods, we have built a detailed and realistic simulator, and
use it to generate data with which our methods can be evaluated. We first describe the simulator
we have built for these purposes. We next present results using the simulator under a number of
different conditions, including complex noise models validated through real-world experiments.
For all results in this section, we use a virtual setup with three base stations, and evaluate on a
0 points canvassing the entire 50x50 virtual space corresponding to a 3m
by 3m grid. All results are reported in terms of Median Square Error (Median SE).
Before describing results, we briefly comment on the latency ramifications of SMLPs. Note that
e SMLP models in this section use 2 layers in the Lower Pair Network (LPN) with 500 and 50
neurons per layer, respectively, and 2 layers in the Upper Connection Network with 200 and 50
neurons per layer, respectively. Further, the number of multiply and addition operations needed
can be found according to the following formula:
Therefore, for the SMLP used in this paper, a total of
multiply and addition operations are required. A commodity accelerator capable of executing
4500 GFLOPs can execute these operations in 0.01 milliseconds, making our method applicable
rios with small latency budgets.
In order to evaluate our proposed SMLP Model, we have created a simulator for generating
training data under a number of different scenarios of interest, including equipment failure and
ased the simulator publicly as a community resource. The code for the
simulator can be found online and is under active development (currently available at
https://github.com/KevinHCChen/wireless-aoa/ ).
The simulator first constructs a configurable virtual space in which the experiment will exist. The
simulator then places mobile nodes within the space, as well as base stations. Given this setup,
the simulator generates the data needed to train the SMLP models. Specifically, based on the
tion of the base stations and the mobile nodes, the simulator generates either time of
arrival data to each of the base stations given a virtual mobile node and its
position, and pairs this with the location of the mobile nodes. This data is then exported such that
it can be used in training the SMLP models.
9, No.3, May 2017
29
base stations pairs in the Lower Pair Network (LPN) and then combines them in the Upper Connection
coordinates of a mobile device.
In order to validate our proposed methods, we have built a detailed and realistic simulator, and
use it to generate data with which our methods can be evaluated. We first describe the simulator
ing the simulator under a number of
world experiments.
For all results in this section, we use a virtual setup with three base stations, and evaluate on a
0 points canvassing the entire 50x50 virtual space corresponding to a 3m
Before describing results, we briefly comment on the latency ramifications of SMLPs. Note that
e SMLP models in this section use 2 layers in the Lower Pair Network (LPN) with 500 and 50
neurons per layer, respectively, and 2 layers in the Upper Connection Network with 200 and 50
ddition operations needed
multiply and addition operations are required. A commodity accelerator capable of executing
4500 GFLOPs can execute these operations in 0.01 milliseconds, making our method applicable
In order to evaluate our proposed SMLP Model, we have created a simulator for generating
training data under a number of different scenarios of interest, including equipment failure and
ased the simulator publicly as a community resource. The code for the
simulator can be found online and is under active development (currently available at
ual space in which the experiment will exist. The
simulator then places mobile nodes within the space, as well as base stations. Given this setup,
the simulator generates the data needed to train the SMLP models. Specifically, based on the
tion of the base stations and the mobile nodes, the simulator generates either time of
arrival data to each of the base stations given a virtual mobile node and its
ta is then exported such that
10. International Journal of Computer Networks & Communications (IJCNC) Vol.
The simulator can be easily configured via initialization files, whose parameters include:
1. Distribution of mobile locations (uniform grid or random)
2. Number of mobile locations
3. Dimensionality of simulation (2D or 3D)
4. Number and location of base stations
In order to simulate realistic conditions that real
several noise and filtering modules to emulate fading and multipath effects.
1. Gaussian noise
2. Angle Dependent Noise
3. Base station outages
4. Uniform noise
We describe these noise models further in Section 5.3 below.
5.2. Collinearity Mitigation
In this section, we demonstrate that SMLPs are able to rectify localizati
regions. As described in Section 3.1, the presence of collinear regions between base stations,
even when a sufficient number of base stations are used, introduce difficulties for accurate
localization of mobile nodes.
We train an SMLP model with 2 layers in the Lower Pair Network (LPN)
neurons per layer, respectively, and 2 layers in the Upper Connection Network with 200 and 50
neurons per layer, respectively. As a point of comparison, we train a conventional
structured) fully connected MLP model with 4 layers and 500, 50, 200, 50 neurons per layer,
respectively. Figure 4 shows the error in localizing points within the simulated space with three
base stations for the MLP and SMLP models, where the error i
predicted and true locations of each mobile node. We find that while the MLP model has large
amounts of error in the collinear regions between base stations (denoted by the red regions in
Figure 4 on the left), the SMLP mod
those in non-collinear regions (as seen in the blue regions in Figure 4 on the right). Note that in
order to make a fairer comparison, we additionally train a not
per layer such that the total number of neurons is comparable to that of the SMLP, and find that
this does not have a marked impact on improving its performance.
Figure 4: When simulating data without noise, the non
collinear regions while our proposed Structured MLP (right) model can.
International Journal of Computer Networks & Communications (IJCNC) Vol.9, No.3, May 2017
The simulator can be easily configured via initialization files, whose parameters include:
Distribution of mobile locations (uniform grid or random)
Number of mobile locations
Dimensionality of simulation (2D or 3D)
Number and location of base stations
In order to simulate realistic conditions that real-world deployments will face, we also incorporate
several noise and filtering modules to emulate fading and multipath effects. These include:
We describe these noise models further in Section 5.3 below.
In this section, we demonstrate that SMLPs are able to rectify localization errors within collinear
regions. As described in Section 3.1, the presence of collinear regions between base stations,
even when a sufficient number of base stations are used, introduce difficulties for accurate
n an SMLP model with 2 layers in the Lower Pair Network (LPN) with 500
neurons per layer, respectively, and 2 layers in the Upper Connection Network with 200 and 50
neurons per layer, respectively. As a point of comparison, we train a conventional
structured) fully connected MLP model with 4 layers and 500, 50, 200, 50 neurons per layer,
respectively. Figure 4 shows the error in localizing points within the simulated space with three
base stations for the MLP and SMLP models, where the error is defined in meters between the
predicted and true locations of each mobile node. We find that while the MLP model has large
amounts of error in the collinear regions between base stations (denoted by the red regions in
Figure 4 on the left), the SMLP model is able to resolve these regions with errors comparable to
collinear regions (as seen in the blue regions in Figure 4 on the right). Note that in
order to make a fairer comparison, we additionally train a not-structured MLP with more neuro
per layer such that the total number of neurons is comparable to that of the SMLP, and find that
this does not have a marked impact on improving its performance.
Figure 4: When simulating data without noise, the non-structured MLP (left) is unable to
collinear regions while our proposed Structured MLP (right) model can.
9, No.3, May 2017
30
The simulator can be easily configured via initialization files, whose parameters include:
world deployments will face, we also incorporate
These include:
on errors within collinear
regions. As described in Section 3.1, the presence of collinear regions between base stations,
even when a sufficient number of base stations are used, introduce difficulties for accurate
with 500 and 50
neurons per layer, respectively, and 2 layers in the Upper Connection Network with 200 and 50
neurons per layer, respectively. As a point of comparison, we train a conventional (not-
structured) fully connected MLP model with 4 layers and 500, 50, 200, 50 neurons per layer,
respectively. Figure 4 shows the error in localizing points within the simulated space with three
s defined in meters between the
predicted and true locations of each mobile node. We find that while the MLP model has large
amounts of error in the collinear regions between base stations (denoted by the red regions in
el is able to resolve these regions with errors comparable to
collinear regions (as seen in the blue regions in Figure 4 on the right). Note that in
structured MLP with more neurons
per layer such that the total number of neurons is comparable to that of the SMLP, and find that
resolve the
11. International Journal of Computer Networks & Communications (IJCNC) Vol.
We next examine the Median SE achieved by each of the models in localizing all 2500 data
points within the simulated space. Figure 5 shows the Median SE achieved by the SM
non-structured MLP models for different amounts of training data, ranging from 50 to 2000
points. We find that for any amount of training data used, the SMLP outperforms the MLP in
terms of Median SE.
Intuitively, this is due to the fact that gi
collinear region, the UCN in the SMLP is able to use the output from the two LCN submodels to
resolve the collinear region problem, while the MLP, even with the same number of layers, is not
able to do so.
Figure 5: Using simulation data without noise, we show the improvement in localization accuracy
Median SE, of the Structured MLP (SMLP) as compared to the non
find that for a given accuracy, the SMLP can
by the non
Further, in real-world deployments in which data must be collected, it is of interest to be able to
train the model with as few training points as possible while retaining its ability to localize a
mobile node to high accuracy, such as sub
amount of error is far less than one meter. Importantly, we find that with as few as 50 training
points, the localization is only 0.40 meters (the Median SE is 0.16 meters).
5.3. Resilience Against Noise
We now present results in the presence of noise. In real
of different forms of noise that will be present in the data. To evaluate the performance of our
proposed localization methods under various amounts of noise in a co
we include several noise models within the simulator and test the models under identical
conditions. We concentrate on two main types of noise, Additive Gaussian White Noise
(AGWN) and Angle Dependent Noise.
The first type of noise, AGWN, represents the type of noise that arises from antenna movement,
inaccuracies in measurement data, equipment mis
AGWN data, we sample n points from a Gaussian distribution with mean
deviation σ, where n is the number of data points in the dataset, and then add it to the generated
input data. In experiments, we use µ
International Journal of Computer Networks & Communications (IJCNC) Vol.9, No.3, May 2017
We next examine the Median SE achieved by each of the models in localizing all 2500 data
points within the simulated space. Figure 5 shows the Median SE achieved by the SM
structured MLP models for different amounts of training data, ranging from 50 to 2000
points. We find that for any amount of training data used, the SMLP outperforms the MLP in
Intuitively, this is due to the fact that given two points in different positions but within the same
collinear region, the UCN in the SMLP is able to use the output from the two LCN submodels to
resolve the collinear region problem, while the MLP, even with the same number of layers, is not
Figure 5: Using simulation data without noise, we show the improvement in localization accuracy
of the Structured MLP (SMLP) as compared to the non-structured MLP model. Further, we
find that for a given accuracy, the SMLP can localize nodes using only a fraction of training data required
by the non-structured MLP for comparable performance.
world deployments in which data must be collected, it is of interest to be able to
train the model with as few training points as possible while retaining its ability to localize a
mobile node to high accuracy, such as sub-meter. At all amounts of training data in Figure 5, the
amount of error is far less than one meter. Importantly, we find that with as few as 50 training
points, the localization is only 0.40 meters (the Median SE is 0.16 meters).
5.3. Resilience Against Noise
present results in the presence of noise. In real-world deployments, there are a number
of different forms of noise that will be present in the data. To evaluate the performance of our
under various amounts of noise in a consistent and repeatable way,
we include several noise models within the simulator and test the models under identical
conditions. We concentrate on two main types of noise, Additive Gaussian White Noise
(AGWN) and Angle Dependent Noise.
of noise, AGWN, represents the type of noise that arises from antenna movement,
inaccuracies in measurement data, equipment mis-calibrations, and interference. To generate the
points from a Gaussian distribution with mean µ and sta
is the number of data points in the dataset, and then add it to the generated
input data. In experiments, we use µ=0 so the noise is centered around the true value and vary
9, No.3, May 2017
31
We next examine the Median SE achieved by each of the models in localizing all 2500 data
points within the simulated space. Figure 5 shows the Median SE achieved by the SMLP and
structured MLP models for different amounts of training data, ranging from 50 to 2000
points. We find that for any amount of training data used, the SMLP outperforms the MLP in
ven two points in different positions but within the same
collinear region, the UCN in the SMLP is able to use the output from the two LCN submodels to
resolve the collinear region problem, while the MLP, even with the same number of layers, is not
Figure 5: Using simulation data without noise, we show the improvement in localization accuracy, in
structured MLP model. Further, we
localize nodes using only a fraction of training data required
world deployments in which data must be collected, it is of interest to be able to
train the model with as few training points as possible while retaining its ability to localize a
mounts of training data in Figure 5, the
amount of error is far less than one meter. Importantly, we find that with as few as 50 training
world deployments, there are a number
of different forms of noise that will be present in the data. To evaluate the performance of our
nsistent and repeatable way,
we include several noise models within the simulator and test the models under identical
conditions. We concentrate on two main types of noise, Additive Gaussian White Noise
of noise, AGWN, represents the type of noise that arises from antenna movement,
calibrations, and interference. To generate the
µ and standard
is the number of data points in the dataset, and then add it to the generated
=0 so the noise is centered around the true value and vary the
12. International Journal of Computer Networks & Communications (IJCNC) Vol.
value of σ.
The second type of noise is Angle De
that measurements become noisier as the angle between the normal from the antenna array on the
base station and the mobile node increases, denoted by
mobile node is aligned in front of the base station antenna (i.e., close to the normal from the
antenna array), there is little noise compared with when the mobile node is oriented, for example,
near 90 degrees from the normal of the antenna array. Accordingl
function to model the effect of the noise, as defined in Equation 1, where angle
in Figure 1, k controls for the magnitude of the nonlinearity, and
noise (i.e., noise when the angle is 0 degrees).
(1) | |
This model reflects what we find through empirical observation in the field with equipment.
In generating data under this noise model, we first generate noise from a Gaussian distribution
and then multiply it by the nonlinear factor
in Equation 2.
(2) Ɲ 0,
Just as in the case of the AGWN model, the Angle Dependent Noise model can be automatically
added and configured with different parameter
described the two noise models, we examine how our data
of noise.
Figure 6: Using simulation data with Additive Gaussian White Noise (AGWN), we show the improvement
in localization accuracy of the Structured MLP (SMLP) as compared to the non
We first examine the impact on performance under the AGWN model. Figure 6 shows the impact
of the AGWN model on both the proposed SMLP method and the non
different amounts of noise and different amounts of training data
data is the number of sampled locations
µ=0 and σ={0.01, 0.02, 0.05}. These standard
angular noise of 2.9 degrees, 5.7 degrees, and 14.4 degrees, respectively. We find that the model
is robust to small to medium amounts of Gaussian noise, both achieving accuracies well within
International Journal of Computer Networks & Communications (IJCNC) Vol.9, No.3, May 2017
The second type of noise is Angle Dependent Noise. The Angle Dependent Noise models the fact
that measurements become noisier as the angle between the normal from the antenna array on the
base station and the mobile node increases, denoted by θ in Figure 1. More specifically, when the
le node is aligned in front of the base station antenna (i.e., close to the normal from the
antenna array), there is little noise compared with when the mobile node is oriented, for example,
near 90 degrees from the normal of the antenna array. Accordingly, we use a nonlinearity
to model the effect of the noise, as defined in Equation 1, where angle θ
controls for the magnitude of the nonlinearity, and j controls for the base amount of
gle is 0 degrees).
This model reflects what we find through empirical observation in the field with equipment.
In generating data under this noise model, we first generate noise from a Gaussian distribution
by the nonlinear factor according to the location of the point, as shown
Just as in the case of the AGWN model, the Angle Dependent Noise model can be automatically
added and configured with different parameters for use by the simulator. Now that we have
described the two noise models, we examine how our data-driven methods operate in the presence
Figure 6: Using simulation data with Additive Gaussian White Noise (AGWN), we show the improvement
in localization accuracy of the Structured MLP (SMLP) as compared to the non-structured MLP model.
We first examine the impact on performance under the AGWN model. Figure 6 shows the impact
of the AGWN model on both the proposed SMLP method and the non-structured MLP for
different amounts of noise and different amounts of training data, where the amount of
data is the number of sampled locations. The noise is sampled from Gaussian distributions with
={0.01, 0.02, 0.05}. These standard deviation settings translate to average amounts of
angular noise of 2.9 degrees, 5.7 degrees, and 14.4 degrees, respectively. We find that the model
is robust to small to medium amounts of Gaussian noise, both achieving accuracies well within
9, No.3, May 2017
32
pendent Noise. The Angle Dependent Noise models the fact
that measurements become noisier as the angle between the normal from the antenna array on the
in Figure 1. More specifically, when the
le node is aligned in front of the base station antenna (i.e., close to the normal from the
antenna array), there is little noise compared with when the mobile node is oriented, for example,
y, we use a nonlinearity
to model the effect of the noise, as defined in Equation 1, where angle θ is defined
controls for the base amount of
This model reflects what we find through empirical observation in the field with equipment.
In generating data under this noise model, we first generate noise from a Gaussian distribution
according to the location of the point, as shown
Just as in the case of the AGWN model, the Angle Dependent Noise model can be automatically
s for use by the simulator. Now that we have
operate in the presence
Figure 6: Using simulation data with Additive Gaussian White Noise (AGWN), we show the improvement
structured MLP model.
We first examine the impact on performance under the AGWN model. Figure 6 shows the impact
tructured MLP for
the amount of training
. The noise is sampled from Gaussian distributions with
deviation settings translate to average amounts of
angular noise of 2.9 degrees, 5.7 degrees, and 14.4 degrees, respectively. We find that the model
is robust to small to medium amounts of Gaussian noise, both achieving accuracies well within
13. International Journal of Computer Networks & Communications (IJCNC) Vol.
the 1 meter fidelity goal. Further, we find that the SMLP significantly outperforms the non
structured MLP at all settings of noise and amounts of training data, with the gap between the
models increasing as σ decreases. Finally, we find that at even high levels of
sufficient amount of training data (250 points), the SMLP model is able to meet the 1 meter
fidelity goal, and with smaller amounts of training data, performance suffers only marginally
(error rising from 0.99 median meters at 250 training p
training points). Importantly, this shows that even under a significant amount of noise, the SMLP
model can still approximately meet the 1 meter fidelity goal with little data.
We next examine how the addition of
Figure 6 shows the impact of Gaussian noise as the number of training samples is varied for
different levels of noise. We find that at small to medium levels of noise, even with as few as 100
points, the 1 meter fidelity goal is easily reached (with a localization error of 0.58 median meters
and 0.75median meters for σ = 0.01 and 0.02, respectively).
Figure 7: Our Angle Dependent noise model
captured from field experiments (points)
model mirrors
We now examine the impact of adding Angle Dependent Noise. For these experiments, we set the
parameters as k=1 and j=4 in the nonlinearity defined in Equation 1. These parameters are set
based on empirical observation of field experiments performed with a
Radio Peripheral (USRP) software defined radios, using the setup as described in Section 6 to
measure the amount of Angle Dependent Noise present in a true system deployment. This allows
us to properly set the parameters in the
Equation 1 with these parameters superimposed on top of real data collected in the field, where
the x-axis shows the true angle offset between the mobile node and the normal from the antenna
array, and the y-axis show the amount of error in measuring the angle
The figure shows that the nonlinearity model
parameter settings.
Given these parameters, we simulate Angle Dependent Noi
noise, σ={0.01, 0.02, 0.05}, multiplied by the nonlinearity defined in Equation 1. Figure 8 shows
the impact of Angle Dependent Noise for different amounts of Gaussian noise and varying
amounts of training data. Overall,
Dependent Noise model is less than that under the AGWN model. This can be explained by the
fact that the Angle Dependent Noise model largely only affects those points at a greater angular
International Journal of Computer Networks & Communications (IJCNC) Vol.9, No.3, May 2017
. Further, we find that the SMLP significantly outperforms the non
structured MLP at all settings of noise and amounts of training data, with the gap between the
σ decreases. Finally, we find that at even high levels of noise, given a
sufficient amount of training data (250 points), the SMLP model is able to meet the 1 meter
fidelity goal, and with smaller amounts of training data, performance suffers only marginally
(error rising from 0.99 median meters at 250 training points to 1.03 median meters with only 100
). Importantly, this shows that even under a significant amount of noise, the SMLP
model can still approximately meet the 1 meter fidelity goal with little data.
We next examine how the addition of noise impacts how much data is needed to train the model.
Figure 6 shows the impact of Gaussian noise as the number of training samples is varied for
different levels of noise. We find that at small to medium levels of noise, even with as few as 100
ts, the 1 meter fidelity goal is easily reached (with a localization error of 0.58 median meters
σ = 0.01 and 0.02, respectively).
oise model with k=1 and j=4 (red dotted line) reflects real
(points) over various angles (x-axis), as shown by the fact that the noise
mirrors the real-world points as the angle varies.
We now examine the impact of adding Angle Dependent Noise. For these experiments, we set the
=4 in the nonlinearity defined in Equation 1. These parameters are set
based on empirical observation of field experiments performed with a set of N200 User Software
Radio Peripheral (USRP) software defined radios, using the setup as described in Section 6 to
measure the amount of Angle Dependent Noise present in a true system deployment. This allows
us to properly set the parameters in the nonlinearity. Figure 7 shows the nonlinearity defined in
Equation 1 with these parameters superimposed on top of real data collected in the field, where
axis shows the true angle offset between the mobile node and the normal from the antenna
axis show the amount of error in measuring the angle-of-arrival to that point.
The figure shows that the nonlinearity model tracks real-world measurements under these
Given these parameters, we simulate Angle Dependent Noise for varying levels of Gaussian
={0.01, 0.02, 0.05}, multiplied by the nonlinearity defined in Equation 1. Figure 8 shows
the impact of Angle Dependent Noise for different amounts of Gaussian noise and varying
amounts of training data. Overall, we find that the amount of localization error under the Angle
Dependent Noise model is less than that under the AGWN model. This can be explained by the
fact that the Angle Dependent Noise model largely only affects those points at a greater angular
9, No.3, May 2017
33
. Further, we find that the SMLP significantly outperforms the non-
structured MLP at all settings of noise and amounts of training data, with the gap between the
noise, given a
sufficient amount of training data (250 points), the SMLP model is able to meet the 1 meter
fidelity goal, and with smaller amounts of training data, performance suffers only marginally
s with only 100
). Importantly, this shows that even under a significant amount of noise, the SMLP
noise impacts how much data is needed to train the model.
Figure 6 shows the impact of Gaussian noise as the number of training samples is varied for
different levels of noise. We find that at small to medium levels of noise, even with as few as 100
ts, the 1 meter fidelity goal is easily reached (with a localization error of 0.58 median meters
reflects real-world data
, as shown by the fact that the noise
We now examine the impact of adding Angle Dependent Noise. For these experiments, we set the
=4 in the nonlinearity defined in Equation 1. These parameters are set
set of N200 User Software
Radio Peripheral (USRP) software defined radios, using the setup as described in Section 6 to
measure the amount of Angle Dependent Noise present in a true system deployment. This allows
nonlinearity. Figure 7 shows the nonlinearity defined in
Equation 1 with these parameters superimposed on top of real data collected in the field, where
axis shows the true angle offset between the mobile node and the normal from the antenna
arrival to that point.
world measurements under these
se for varying levels of Gaussian
={0.01, 0.02, 0.05}, multiplied by the nonlinearity defined in Equation 1. Figure 8 shows
the impact of Angle Dependent Noise for different amounts of Gaussian noise and varying
we find that the amount of localization error under the Angle
Dependent Noise model is less than that under the AGWN model. This can be explained by the
fact that the Angle Dependent Noise model largely only affects those points at a greater angular
14. International Journal of Computer Networks & Communications (IJCNC) Vol.
offset from the normal from the antenna and the AGWN model affects all points equally. We find
that the model is able to achieve the goal of at least 1 meter fidelity with small, medium, and large
amounts of Gaussian noise under the Angle Dependent Noise model
significantly outperforms the MLP with small and medium amounts of noise at all amounts of
training data, with the gap between the models again increasing as
SMLP model at σ=0.02 is able to achieve appr
MLP model at σ=0.01 with small amounts of training data. This shows that under the Angle
Dependent Noise model, the SMLP model can handle up to twice the amount of noise as the MLP
model.
We also examine how the addition of noise impacts how much data is needed to train the model.
Figure 8 shows the impact of Gaussian noise under the angle dependent model as the number of
training samples is varied for different levels of noise. We find that at all levels
noise, even with as few as 100 points, the 1 meter fidelity goal is easily reached (with a
localization error of 0.51m, 0.65m, and 0.88m, for
within the 1 meter goal).
Figure 8: Using simulation data with Angle Dependent Noise, we show the improvement in localization
accuracy of the Structured MLP (SMLP) as compared to the non
5.4. 3D Simulations
For completeness, we also include a 3D component to our simulator in order to verify our models
with the added dimensionality. In practice, to capture a 3D position, the base stations must have
antennas in two polarities, usually vertical and horizontal. O
within the 3D space, and then calculates angle
this, we can evaluate our models for 3D data in the same manner as we do with 2D data.
We find that the 3D results follow
is due to the fact that the primary difference between MLP and SMLP models is their ability to
address the collinear region, which remains roughly the same in size in both the 2D and 3D case
as the region lies within the 2D plane with minimal impact in the added third dimension. Figure 9
is a 3D rendering of the predicted positions and associated Median SE from two different
viewpoints: a top-corner view and a ceiling view. As previously s
International Journal of Computer Networks & Communications (IJCNC) Vol.9, No.3, May 2017
et from the normal from the antenna and the AGWN model affects all points equally. We find
that the model is able to achieve the goal of at least 1 meter fidelity with small, medium, and large
amounts of Gaussian noise under the Angle Dependent Noise model. Further, the SMLP
significantly outperforms the MLP with small and medium amounts of noise at all amounts of
training data, with the gap between the models again increasing as σ decreases. Interestingly, the
=0.02 is able to achieve approximately the same error as the non
=0.01 with small amounts of training data. This shows that under the Angle
Dependent Noise model, the SMLP model can handle up to twice the amount of noise as the MLP
ow the addition of noise impacts how much data is needed to train the model.
Figure 8 shows the impact of Gaussian noise under the angle dependent model as the number of
training samples is varied for different levels of noise. We find that at all levels
noise, even with as few as 100 points, the 1 meter fidelity goal is easily reached (with a
localization error of 0.51m, 0.65m, and 0.88m, for σ = 0.01, 0.02, and 0.05, respectively, all well
ation data with Angle Dependent Noise, we show the improvement in localization
accuracy of the Structured MLP (SMLP) as compared to the non-structured MLP model.
For completeness, we also include a 3D component to our simulator in order to verify our models
added dimensionality. In practice, to capture a 3D position, the base stations must have
antennas in two polarities, usually vertical and horizontal. Our simulator generates positions
within the 3D space, and then calculates angle-of-arrival or time-of-flight for each polarity. From
this, we can evaluate our models for 3D data in the same manner as we do with 2D data.
We find that the 3D results follow the same trends presented in the previous results section. This
is due to the fact that the primary difference between MLP and SMLP models is their ability to
address the collinear region, which remains roughly the same in size in both the 2D and 3D case
as the region lies within the 2D plane with minimal impact in the added third dimension. Figure 9
is a 3D rendering of the predicted positions and associated Median SE from two different
corner view and a ceiling view. As previously seen for the 2D results, the error
9, No.3, May 2017
34
et from the normal from the antenna and the AGWN model affects all points equally. We find
that the model is able to achieve the goal of at least 1 meter fidelity with small, medium, and large
. Further, the SMLP
significantly outperforms the MLP with small and medium amounts of noise at all amounts of
decreases. Interestingly, the
oximately the same error as the non-structured
=0.01 with small amounts of training data. This shows that under the Angle
Dependent Noise model, the SMLP model can handle up to twice the amount of noise as the MLP
ow the addition of noise impacts how much data is needed to train the model.
Figure 8 shows the impact of Gaussian noise under the angle dependent model as the number of
of simulated
noise, even with as few as 100 points, the 1 meter fidelity goal is easily reached (with a
= 0.01, 0.02, and 0.05, respectively, all well
ation data with Angle Dependent Noise, we show the improvement in localization
structured MLP model.
For completeness, we also include a 3D component to our simulator in order to verify our models
added dimensionality. In practice, to capture a 3D position, the base stations must have
ur simulator generates positions
flight for each polarity. From
this, we can evaluate our models for 3D data in the same manner as we do with 2D data.
the same trends presented in the previous results section. This
is due to the fact that the primary difference between MLP and SMLP models is their ability to
address the collinear region, which remains roughly the same in size in both the 2D and 3D cases,
as the region lies within the 2D plane with minimal impact in the added third dimension. Figure 9
is a 3D rendering of the predicted positions and associated Median SE from two different
een for the 2D results, the error
15. International Journal of Computer Networks & Communications (IJCNC) Vol.9, No.3, May 2017
35
concentrates itself along the collinear region along the straight line between base stations. We
have omitted complete results for the 3D simulations due to space concerns, as they very closely
mirror that of the 2D results.
Figure 9: Our simulator is capable of generating and testing our localization methods under 2D and 3D
environments, as shown in the two 3D views above.
6. REAL-WORLD EXPERIMENTS
In addition to the simulated experiments discussed in the previous section, in this section we
present a real-world experiment performed in a realistic deployment scenario. We first describe
the experimental equipment. Our experimental setup consists of a collection of receivers, a
transmitter, and a reference node. We use N200 User Software Radio Peripheral (USRP) software
defined radios operating at 916 MHz. The receiver USPRs are connected to a portable Dell
Inspiron 410 via a Netgear gigabit switch, and serves as a control node responsible for starting
experiments and collecting measurements from the USRPs. The transmitter sends a predefined
symbol string of a thousand 1s that is captured by all of the receiver USRPs. The reference point
transmits the same symbol string and is placed directly in front of the base station so the phase
offsets can be used to calibrate the transmitter values. We follow the same experimental approach
and reference node implementation described in [15].
16. International Journal of Computer Networks & Communications (IJCNC) Vol.
Figure 10: Layout of indoor and outdoor experimental setups: a five by six grid for t
location (for the mobile node)
In order to evaluate our models on real
use a five by six grid of location points, resulting
station locations. However, due to equipment limitations (having only one base station node
consisting of four USRPs as an antenna array of four antennas
three locations sequentially and re
having three base stations measuring at once. Note that we use a two
lack an antenna array with vertical polarity. However, based on the similarity between
3D simulation results, we expect that 3D results in the field would be very similar to the results
presented here. The environment is an outside, open, grassy area, as shown in Figure 11.
Figure 11: The real
For the experiment, we collect 90 samples (3 samples at each of the 30 points on the grid) per day
for two days. The model is trained on data from one day, and tested on data from the other day,
International Journal of Computer Networks & Communications (IJCNC) Vol.9, No.3, May 2017
Figure 10: Layout of indoor and outdoor experimental setups: a five by six grid for the transmitting USRP
(for the mobile node) with three base station (BS) locations around the grid
In order to evaluate our models on real-world datasets, we use the setup shown in Figure 10. We
use a five by six grid of location points, resulting in 30 locations in total. We use three base
station locations. However, due to equipment limitations (having only one base station node
as an antenna array of four antennas), we move the base station to the
ially and re-measure at each location, but logically this is equivalent to
having three base stations measuring at once. Note that we use a two-dimensional setup, as we
lack an antenna array with vertical polarity. However, based on the similarity between
3D simulation results, we expect that 3D results in the field would be very similar to the results
presented here. The environment is an outside, open, grassy area, as shown in Figure 11.
Figure 11: The real-world experiment environment
For the experiment, we collect 90 samples (3 samples at each of the 30 points on the grid) per day
for two days. The model is trained on data from one day, and tested on data from the other day,
9, No.3, May 2017
36
he transmitting USRP
locations around the grid
world datasets, we use the setup shown in Figure 10. We
in 30 locations in total. We use three base
station locations. However, due to equipment limitations (having only one base station node
), we move the base station to the
measure at each location, but logically this is equivalent to
dimensional setup, as we
lack an antenna array with vertical polarity. However, based on the similarity between our 2D and
3D simulation results, we expect that 3D results in the field would be very similar to the results
presented here. The environment is an outside, open, grassy area, as shown in Figure 11.
For the experiment, we collect 90 samples (3 samples at each of the 30 points on the grid) per day
for two days. The model is trained on data from one day, and tested on data from the other day,
17. International Journal of Computer Networks & Communications (IJCNC) Vol.9, No.3, May 2017
37
where the second day is completely held out from training (i.e., the model is never trained and
tested on data collected on the same day). We find that the SMLP achieves a Median SE of 0.16
meters (0.4 meter error) in this experiment, achieving sub-meter accuracy, while the unstructured
DNN performs worse, and achieves a Median SE of 0.2 meters (0.45 meter error).
7. IMPACT
In this section, we describe the impact of our work presented in this paper. As shown in Sections
5 and 6, our proposed methods are able to achieve sub-meter localization accuracy for mobile
nodes, even in the presence of noise. mmWave transmitters and receivers will likely need to
align their beams via a beam forming process, as the alignment accuracy between the transmitter
and receiver will determine the quality of the signal between the two. However, the exact impact
of localization accuracy on signal quality, as well as baseline levels of localization accuracy
needed for high-rate mmWave communication, is dependent on antenna choices, specifically the
choice of omni or directional antennas on base stations and mobile nodes, and the half power
beam width of the antennas. As such, we have set the 1 meter indoor localization fidelity goal in
the 5G standard [2] as our goal in developing the methods. In this section, we go beyond this
exact numerical goal, and describe the potential impact of our method by considering two
possible systems and the ramifications of our localization on these systems.
Specifically, we consider setups in which the antenna and equipment choices result in 1) the
existence of an exact beam alignment that allows for higher signal quality than any other
alignment choice (i.e., there exists one alignment, such as an exact alignment between the
transmitter and receiver, that is strictly better than all other alignments), and 2) the existence of a
window of alignments such that as long as the alignment of the transmitter and receiver is within
this window, signal quality is identical or near identical (i.e., for a given distance, as long as the
alignment is within some x amount of error from the true alignment, there is no impact on signal
quality, where x will decrease as distance between the transmitter and receiver increases). Note
that scenario one would arise when using highly directional antennas with small half power beam
widths, while scenario two would arise when using an omni-directional antenna for either the
transmitter or receiver, or using directional antennas with relatively large half power beam
widths.
Under scenario one, our improved localization accuracy will strictly improve the signal quality
between the transmitter and receiver. As there exists an optimal alignment between the
transmitter and receiver, which will commonly correspond to the line-of-sight alignment between
the transmitter and receiver, improved localization accuracy will allow for an as near to optimal
alignment as is possible, subject only to localization error. As such, under this scenario, our
improved localization accuracy will improve signal quality as compared with less accurate
methods.
Under scenario two, for a given distance between a transmitter and receiver, our improved
localization method will only improve upon other methods in situations in which other
localization methods are not sufficiently accurate as to allow for alignment within the error
window. Beyond this however, our method does offer significant gains under scenario two when
the distance between the transmitter and receiver increases. As this distance between the
transmitter and receiver increases, the size of the error window will decrease. In this respect, the
more accurate localization offered by our method will allow for greater distances between the
transmitter and receiver to be used without affecting signal quality due to misalignment.
(However, signal attenuation and other effects may still impact performance as distance
increases).
18. International Journal of Computer Networks & Communications (IJCNC) Vol.9, No.3, May 2017
38
8. CONCLUSION
In this paper, we have introduced Structured Multilayer Perceptrons (SMLPs), a data-driven deep
neural network (DNN) localization approach to enable precise narrow beam alignment required
by mmWave to deliver data at 10Gb/s or higher rates. Our indoor simulations and real world
experiment outdoors demonstrate that we can achieve sub-meter localization accuracy. In
general, the localization capabilities of this paper will allow other layers in the network stack to
provide improved access control, and therefore improve overall system throughput. Future work
will focus on more real-world experiments using these low frequency localization methods, as
well as integrating these capabilities into the networking stack in order to take advantage of
accurate location information to power smarter, faster next-generation wireless networks.
ACKNOWLEDGEMENTS
This work is supported in part by gifts from the Intel Corporation and in part by the Naval Supply
Systems Command award under the Naval Postgraduate School Agreements No. N00244-15-
0050 and No. N00244-16-1-0018.
REFERENCES
[1] T. Nitsche, C. Cordeiro, A. B. Flores, E. W. Knightly, E. Perahia, and J. C. Widmer,“IEEE
802.11ad: directional 60 ghz communication for multi-gigabit-per-second Wi-Fi [invited
paper],”IEEE Communications Magazine, vol. 52, no. 12, pp. 132–141, 2014.
[2] N. Alliance, “5g white paper,”Next Generation Mobile Networks, White paper, 2015.
[3] R, Schmidt. "Multiple emitter location and signal parameter estimation," IEEE transactions on
antennas and propagation vo. 3, no. 3, pp. 276-280, 1986.
[4] F. Adib, Z. Kabelac, and D. Katabi, “Multi-person localization via rf body reflections,” in12th
USENIX Symposium on Networked Systems Design and Implementation (NSDI 15), pp. 279–
292, 2015.
[5] F. Adib, Z. Kabelac, D. Katabi, and R. C. Miller, “3d tracking via body radio reflections,” in
NSDI, vol. 14, pp. 317–329, 2014.
[6] I. . working group et al., “Ieee 802.11ad, amendment 3: Enhancements for very high throughput in
the 60 ghz band,” 2012.
[7] T. Nitsche, A. B. Flores, E. W. Knightly, and J. Widmer, “Steering with eyes closed:mm-wave
beam steering without in-band measurement,” in2015 IEEE Conference on Computer
Communications (INFOCOM), pp. 2416–2424, IEEE, 2015.
[8] J. Xiong and K. Jamieson, “Arraytrack: A fine-grained indoor location system,”in NSDI, pp. 71–
84, 2013.
[9] G. Félix, M. Siller, and E. N. Álvarez. "A fingerprinting indoor localization algorithm based deep
learning," Ubiquitous and Future Networks (ICUFN), 2016 Eighth International Conference on.
IEEE, 2016.
[10] X. Wang, L. Gao, S. Mao, and S. Pandey. "CSI-based fingerprinting for indoor localization: A
deep learning approach," IEEE Transactions on Vehicular Technology vol 66, no. 1, pp. 763-776,
2017.
[11] J. Luo and H. Gao. "Deep belief networks for fingerprinting indoor localization using
ultrawideband technology," International Journal of Distributed Sensor Networks, 2016.
[12] S. J. Tarsa, M. Comiter, M. B. Crouse, B. McDanel, and H. Kung, “Taming wireless fluctuations
by predictive queuing using a sparse-coding link-state model,” in Proceedings of the 16th ACM
International Symposium on Mobile Ad Hoc Networkingand Computing, pp. 287–296, ACM,
2015.
[13] M. Courbariaux, Y. Bengio, and J.-P. David, “Binary connect: Training deep neural networks with
binary weights during propagations,” in Advances in Neural Information Processing Systems, pp.
3123–3131, 2015.
[14] B. McDanel, S. Teerapittayanon, and H. Kung, “Embedded binarized neural networks,” in 2017
International Conference on Embedded Wireless Systems and Networks (EWSN 2017), 2017.
19. International Journal of Computer Networks & Communications (IJCNC) Vol.
[15] H.C. Chen, T.H. Lin, H. Kung, C.K. Lin, and Y. Gwon, “Determining rf angle of arriva
antenna arrays: a field evaluation,” in
2012, pp. 1–6, IEEE, 2012.
[16] R. M. Vaghefi, R. G. Mohammad, R. M. Buehrer, and E. G. Strom, “Cooperative received signal
strength-based sensor localizati
Signal Processing 61, no. 6, pp. 1389
[17] M. E. Rida, F. Liu, Y. Jadi, A. A. Algawhari, and A. Askourih, “Indoor location position bas
bluetooth signal strength," in Information Science and Control Engineering (I
International Conference 2015, pp. 769
[18] S. He and S-H. G. Chan, “Sectjunction: Wi
sectors,” in Communications (ICC), 2014 IEEE International Conference, pp. 2605
2014.
Authors
Marcus Comiter is a Ph.D. candidate in Computer Science at Harvard
University. He received his A.B. in Computer Science and Statistics
from Harvard University.
Michael Crouse is a Ph.D. candidate in Computer Science at Harvard
University. He received his B.S. and M.S. in Computer Science from
Wake Forest University.
HT Kung is the William H. Gates Professor of Computer Science and
Electrical Engineering at Harvard University. He is interested in
computing, communications and sensing. Prior to joining Harvard in
1992, he taught at Carnegie Mellon for 19 years after receiving his Ph.D.
there. Professor Kung is best known for his pioneering work on I/O
complexity in computing theory, systolic arrays in parallel processing,
and optimistic concurrency control in database systems. His academic
honors include Member of National Academy of Engineering and the
ACM SIGOPS 2015 Hall of Fame Award (with John Robinson) t
recognizes the most influential Operating Systems papers that were
published at least ten years in the past.
International Journal of Computer Networks & Communications (IJCNC) Vol.9, No.3, May 2017
H.C. Chen, T.H. Lin, H. Kung, C.K. Lin, and Y. Gwon, “Determining rf angle of arriva
eld evaluation,” in MilitaryCommunications Conference, 2012-MILCOM
6, IEEE, 2012.
R. M. Vaghefi, R. G. Mohammad, R. M. Buehrer, and E. G. Strom, “Cooperative received signal
based sensor localization with unknown transmit powers,” in IEEE Transactions on
Signal Processing 61, no. 6, pp. 1389-1403, 2013.
M. E. Rida, F. Liu, Y. Jadi, A. A. Algawhari, and A. Askourih, “Indoor location position bas
" in Information Science and Control Engineering (ICISCE) 2nd
International Conference 2015, pp. 769-773, IEEE, 2015.
H. G. Chan, “Sectjunction: Wi-Fi indoor localization based on junction of signal
” in Communications (ICC), 2014 IEEE International Conference, pp. 2605-2610, IEEE,
Marcus Comiter is a Ph.D. candidate in Computer Science at Harvard
University. He received his A.B. in Computer Science and Statistics
Michael Crouse is a Ph.D. candidate in Computer Science at Harvard
University. He received his B.S. and M.S. in Computer Science from
HT Kung is the William H. Gates Professor of Computer Science and
Harvard University. He is interested in
computing, communications and sensing. Prior to joining Harvard in
1992, he taught at Carnegie Mellon for 19 years after receiving his Ph.D.
there. Professor Kung is best known for his pioneering work on I/O
ity in computing theory, systolic arrays in parallel processing,
and optimistic concurrency control in database systems. His academic
honors include Member of National Academy of Engineering and the
ACM SIGOPS 2015 Hall of Fame Award (with John Robinson) that
recognizes the most influential Operating Systems papers that were
published at least ten years in the past.
9, No.3, May 2017
39
H.C. Chen, T.H. Lin, H. Kung, C.K. Lin, and Y. Gwon, “Determining rf angle of arrival using cots
MILCOM
R. M. Vaghefi, R. G. Mohammad, R. M. Buehrer, and E. G. Strom, “Cooperative received signal
” in IEEE Transactions on
M. E. Rida, F. Liu, Y. Jadi, A. A. Algawhari, and A. Askourih, “Indoor location position based on
CISCE) 2nd
d on junction of signal
-2610, IEEE,