This document discusses IoT data processing topologies and considerations. It begins by explaining the types of structured and unstructured data in IoT. It then discusses the importance of processing based on urgency and describes on-site, remote, and collaborative processing topologies. The document also covers IoT device design factors and processing offloading considerations including location, decision making, and other criteria.

1. Ch 6. IoT Processing
Topologies and Type
Dr Loganathan R
1
Introduction to IoT

2. 1 Data Format
Dr Loganathan R 2
The various data generating and
storage sources connected to
the Internet and the plethora of
data types contained within it

3. 1 Data Format
• Data is composed of a variety of data such as e-mails, text documents (Word docs,
PDFs, and others), social media posts, videos, audio files, and images
• Data grouped based on how they can be accessed and stored:
1. Structured data 2. Unstructured data
1.1 Structured data
• Have a pre-defined structure
• Associated with relational database management systems (RDBMS)
• Created by using length-limited data fields such as phone#, SSN, etc.
• Easily searchable by querying algorithms and queries
• Usage: flight or train reservation systems, banking systems, inventory controls
• SQL used for accessing these data in RDBMS
• In IOT Structured data holds a minor share of the total generated data in the Internet.
Dr Loganathan R 3

4. 1 Data Format
1.2 Unstructured data
• All the data on the Internet, which is not structured, is categorized as
unstructured
• No pre-defined structure and can vary according to applications and
data-generating sources
• Examples: text, e-mails, videos, images, phone recordings, chats, etc
• Machine-generated unstructured data: sensor data from traffic,
buildings, industries, satellite imagery, surveillance videos, and others
• Very difficult for querying algorithms to perform a look-up
• Querying languages like NoSQL are used.
Dr Loganathan R 4

5. 2 Importance of Processing in IoT
• Vast amount and types: intelligent and resourceful processing
techniques
• Important: When to process and what to process?
• 3 types based on the urgency of processing
1) Very time critical 2) Time critical 3) Normal
• Very time critical: Flight control systems, healthcare, etc need
immediate decision support are very critical requires very low
threshold of processing latency
• Processing requirements are exceptionally high
• Processing the data in place or almost nearer to the source is crucial
Dr Loganathan R 5

6. 2 Importance of Processing in IoT
• Time critical: Vehicles, traffic, smart home , surveillance , etc, can
tolerate a latency of a few seconds
• Processing requirements allow for the transmission of data to be
processed to remote locations/processors
• Normal: Agriculture, environmental monitoring, etc, can tolerate a
processing latency of a few minutes to a few hours
• No particular time requirements for processing urgently
Dr Loganathan R 6

7. 3 Processing Topologies
• The identification and selection of processing requirement of an IoT
application are the crucial steps in deciding the architecture of the
deployment.
• IoT architecture
• Savings in network bandwidth
• Conserve energy
• Provide allowable processing latencies
• Processing solutions are divided into two large topologies:
1. On-site
2. Off-site.
• The off-site is further divided into:
1. Remote processing
2. Collaborative processing. Dr Loganathan R 7

8. 3.1 On-site processing
Dr Loganathan R 8
• Data is processed at the source itself
• Applications that has a very low tolerance for latencies
• Latency from processing hardware or the network
• Rapid temporal changes can be missed unless the processing
infrastructure is fast and robust enough to handle such data.
• The sensor node processes the information from the sensed event and
generates an alert
Event detection using an
on-site processing topology

9. 3.2 Off-site processing
• Allows for latencies (due to processing or network latencies)
• Cheaper than on-site processing topologies
• The sensor node is responsible for the collection and framing of data to be
transmitted for processing.
• Simpler sensor nodes borrows processing from high-processing enabled
devices to accomplish their tasks
• Multiple nodes can come together to share their processing power in
order to collaboratively process the data
Dr Loganathan R 9

10. 3.2 Off-site processing: Remote processing
• Encompasses sensing data from sensor nodes and then forwards to a
remote server or a cloud-based infrastructure for further processing
and analytics.
• Massive cost and energy savings by the reuse and reallocation of the same
processing resource
• Ensures massive scalability of solutions
• Sensing is local and the decision making is to a remote processor
Dr Loganathan R
10
Event detection using an off-site remote processing topology

11. 3.2 Off-site processing: Collaborative processing
Dr Loganathan R
11
• Limited or no network connectivity
• Economical for large-scale deployments spread over vast areas
• Clubs together the processing power of nearby processing nodes and
collaboratively process the data in the data source itself
• Reduces latencies due to the transfer of data
• Mesh networks for easy implementation
Event detection using a collaborative processing topology

12. 4 IoT Device Design and Selection Considerations
Dr Loganathan R
12
• The processor is the main factor for the IoT device design and
selection for various applications
• Other important considerations are:
• Size: larger the form factor, larger is the energy consumption
• Energy: Higher the energy requirements, higher the energy source (battery)
replacement frequency
• Cost: Cheaper cost of the hardware enables higher density of hardware
deployment
• Memory: volatile and non-volatile memory, higher memory tend to be costlier
• Processing power: decided on what type of sensors can be accommodated with
the IoT device/node and what processing features can integrate on-site
• I/O rating: deciding factor in determining the circuit complexity, energy usage and
requirements for support of various sensing solutions and sensor types
• Add-ons: ADC units, in-built clock circuits, connections to USB and ethernet, inbuilt
wireless access capabilities

13. Dr Loganathan R
5 Processing Offloading
13
• For off-site processing, data from
the sensing layer is forwarded to
the fog or cloud or contained
within the edge layer
• The edge layer use of devices
within the local network to process
data, similar to the collaborative
processing topology
• Fog nodes are localized within a
geographic area and serve the IoT
nodes within a smaller coverage
area
• Forwarding data to a cloud or a
remote server, requires the devices
to be connected to the Internet
The various data generating and storage
sources connected to the Internet and the
plethora of data types contained within it

14. 5 Processing Offloading
Dr Loganathan R
14
• Data offloading is divided into three parts:
1. Offload location (which outlines where all the processing can be offloaded in the
IoT architecture)
2. Offload decision making (how to choose where to offload the processing to and
by how much)
3. Offloading considerations (deciding when to offload).

15. 5.1 Offload location
Dr Loganathan R
15
• The choice of offload location decides the applicability, cost, and
sustainability of the IoT application and deployment
• Edge: Offloading processing to the edge implies that the data processing
is facilitated to a location at or near the source
• Fog: The data, computing, storage and applications are shifted to a place
between the data source and the cloud
• Remote Server: A remote server with good processing power may be
used with IoT-based applications to offload the processing from
resource constrained IoT devices.
• Cloud: A cloud is provisioned for processing offloading so that processing
resources can be rapidly provisioned over the Internet, which can be
accessed globally and enables massive scalability on-demand

16. 5.2 Offload Decision making
Dr Loganathan R
16
• The choice of where to offload and how much to offload is addressed
considering data generation rate, network bandwidth, the criticality of
applications, processing resource available at the offload site and other
factors
• Naive Approach: This rule-based approach, the data from IoT devices are
offloaded to the nearest location based on the achievement of offload
criteria. Statistical measures are consulted for generating the rules
• Bargaining based approach: To maximize the QoS (bandwidth, latencies) by
reach a point where the qualities of certain parameters are reduced, while
the others are enhanced. QoS is achieved collaboratively better for the full
implementation.
• Learning based approach: Rely on past behavior and trends of data flow
through the IoT architecture. The optimization of QoS parameters is done by
learning from historical trends and optimize previous solutions further and
enhance the collective behavior of the IoT implementation

17. 5.3 Offloading considerations
Dr Loganathan R
17
• Offloading parameters need to be considered while deciding upon the
offloading type to choose arise from the nature of the IoT application and
the hardware being used to interact with the application
• Bandwidth: The maximum amount of data that can be simultaneously
transmitted over the network between two points
• Latency: The time delay between the start and completion of an operation,
which is due to physical limitations of the infrastructure in the network or
the processor
• Criticality: Importance of a task being pursued by an IoT application. The
more critical a task is, the lesser latency is expected.
• Resources: It signifies the actual capabilities (like processing power, the
analytical algorithms, etc) of an offload location
• Data volume: The amount of data generated by a source or sources that can
be simultaneously handled by the offload location