An Adaptive Energy Efficient Reliable Routing Protocol for Wireless Sensor Ne...
Energy consumption mitigation routing protocols for large wsn's
1. Energy consumption mitigation Routing Protocols for Large-Scale Wireless Sensor Networks
Anil Kumar H1 ,Manjunath CR2 , Dr Nagaraj GS3
1,2
Dept of CSE,SET,Jain University, 3Prof,Dept of CSE,RVCE,VTU
hmsanilkumar@gmail.com, manjucr123@gmail.com, nagarajgs@yahoo.com
Abstract: With the advances in micro-electronics, II-Energy consumption mitigation-based
wireless sensor devices have been made much smaller category:
and more integrated, and large-scale wireless sensor The routing protocols in this class aim to
networks (WSNs) based the cooperation among the mitigate the energy consumption. They exploit various
significant amount of nodes have become a hot topic. means to achieve this target, such as dynamic event
“Large-scale” means mainly large area or high density of clustering, multi-hop communication, cooperative
a network. Accordingly the routing protocols must scale communication and so on. These methods can consume
well to the network scope extension and node density the energy appropriately and avoid wasted energy [1].
increases. A sensor node is normally energy-limited and
cannot be recharged, and thus its energy consumption has III - Data Gathering algorithm based on Mobile
a quite significant effect on the scalability of the Agent (DGMA)[3]
protocol. In a hierarchical routing protocol, all the nodes
In terms of energy consumption reduction and
are divided into several groups with different assignment
network end-to-end delay decrease, a Data Gathering
levels. The nodes within the high level are responsible
algorithm based on Mobile Agent (DGMA) is proposed
for data aggregation and management work, and the low
for the cluster-based wireless sensor network. where an
level nodes for sensing their surroundings and collecting
emergent event occurs is clustered dynamically based on
information. With focus on the hierarchical structure, in
the event severity, by which the scale and lifetime of
this paper we provide an insight into Energy
clusters are determined. In each cluster a mobile agent is
consumption mitigation routing protocols designed
utilized to traverse every member node to collect sensed
specifically for large-scale WSNs.
data. In the higher level of the network, a virtual cluster
According to the different objectives, the protocols are
is formed among the cluster heads and the base station,
generally classified based on different criteria such as
and multi-hop communication is adopted for sensed data
control overhead reduction, energy consumption
delivery to the base station (BS).
mitigation and energy balance. This paper focuses on the
In DGMA, all the sensor nodes are in “restraining” state
study of energy consumption mitigation to show how to
and they are activated only when some emergent event
mitigate the energy consumption.
occurs. Then the nodes having monitored the event are
clustered. After the event intension gets reduced, the
Keywords: large-scale wireless sensor networks, routing
clustered nodes will change to a “restraining” state for
protocol.
the sake of energy consumption reduction. In the cluster,
the tree structure is used to save energy instead of single
hop communication between the sensor nodes and the
I- Introduction cluster head. After the cluster construction is complete, a
WSN is widely considered as one of the most route for the mobile agent, which is equipped on the
important technologies for the twenty-first century. A cluster head, is used to traverse all the member nodes for
WSN typically consists of a large number of low-cost, collecting the sensed event data. This process is started
low-power, and multifunctional wireless sensor nodes, up by the cluster head and repeated at every cluster
with sensing, wireless communications and computation member by broadcasting a request packet, and
capabilities . These sensor nodes communicate over short anticipating a reply from its each neighbor for getting
distance via a wireless medium and collaborate to residual energy, path loss, and event intension
accomplish a common task, for example, environment information of the neighbor. To deliver the sensed data to
monitoring, military surveillance, and industrial process the final destination (here the base station) in the higher
control.In many WSN applications, the deployment of1 level of the network a virtual cluster is formed wherein
sensor nodes is performed in an ad hoc fashion without the base station acts as the cluster head. As in the local
careful planning and engineering. Once deployed, the cluster, a multi-hop communication is adopted. The
sensor nodes must be able to autonomously organize current cluster head will select the node which is the
themselves into a wireless communication network. closest to the base station in the neighboring nodes as its
Sensor nodes are battery-powered and are expected to next hop. If the distance from all neighbor nodes to the
operate without attendance for a relatively long period of base station is longer than that from the node itself, the
time. In most cases it is very difficult and even node will communicate with the base station directly.
impossible to change or recharge batteries for the sensor When the number of the sensor nodes increases, the
nodes. When the energy of a sensor reaches a certain energy consumption in DGMA increases more slowly.
threshold, the sensor will become faulty and will not be Furthermore, the dynamic cluster formation feature
able to function properly, which will have a major impact further reduces the energy consumption. The use of a
on the network performance [1, 2]. mobile agent reduces energy consumption, but extends
The routing protocols for large scale WSNs can the delay for the cluster head to collect all the sensed data
categorized as control from all the member nodes. The chain-like route delivery
overhead reduction-based, of data by the cluster head makes the node closest to the
energy consumption mitigation-based and base station overloaded and destroys the reliability.
Energy balance-based. Cluster-based wireless sensor network saves energy by
reducing the number of nodes communicating with base
2. station. Compared to direct communication, cluster- Clustering Protocol )BCDCP, the CHs are connected by
based method has a remarkable improving in energy- a tree instead of a club and the BS functions as the
efficient. manager of the whole network, so BCDCP is more
DGMA includes dynamic clustering and Data Gathering energy-efficient than LEACH. DMSTRP improves
Based on Mobile Agent for Emergent Event Monitoring BCDCP further by connecting nodes in clusters by
MSTs. In each cluster, all the nodes including the CH are
Dynamic Clustering connected by a MST and then the CH acts as the leader
a) Dynamic Clustering Based on Event Severity to collect data from the nodes on the tree. On the higher
Degree: level, all the CHs connected by another MST cooperate
After wireless sensor network is deployed into the to route data towards the BS. The data fusion process is
monitoring environment, all nodes will be set to handled during the packet transmission along the tree
“restraining” state rather than clustered. And they’re route.
activated just when some emergent event occurs. Then Obviously, DMSTRP consumes energy more efficiently
the nodes will be clustered. The scale and lifetime of the than LEACH and BCDCP, because the average
clusters lie on the event severity degree. After the transmission distance between nodes is reduced through
stimulating intension is reduced, those activated nodes the multi-hop intra-cluster and inter-cluster
will change to “restraining” state over again. The cluster- communications, and thus the energy dissipation of
tree structure is used to save energy , with multi-hop transmitting data is potentially reduced. Furthermore, due
rather than single hop from the member nodes to the to the reasonable schedule, the transmission collision is
cluster head. alleviated and DMSTRP can achieve shorter delay
compared with LEACH and BCDCP. But the
transmission schedule creates more overhead.
b) The Construction of Virtual Cluster: Generally,
single-hop communication is taken between the cluster V - Hierarchical Geographic Multicast Routing
heads and the base station in spite of long distance, in (HGMR).[5]
which those cluster heads away from the base station HGMR aims at enhancing data forwarding efficiency and
always have a weak lifetime because of more energy increasing the scalability to a large-scale network.
consumption led by long-distance. A multi-hop virtual HGMR seamlessly incorporates the key design concepts
cluster is formed with base station as the cluster head. of the Geographic Multicast Routing (GMR) and
The path from the cluster head to base station can be Hierarchical Rendezvous Point Multicast (HRPM)
searched as follows. The cluster heads always select the protocols, and optimizes the two routing protocols in the
node which is the closest to base station in the neighbor wireless sensor network environment. HGMR starts with
nodes as its next hop. If the distance from all neighbor a hierarchical decomposition of a multicast group into
nodes to base station is longer than that from the node subgroup of manageable size using HRPM’s key concept
itself to base station, the node will communicate with of mobile geographic hashing. Within each subgroup,
base station directly. HGMR uses GMR’s local multicast scheme to forward a
data packet along multiple branches of the multicast tree
Data Gathering Based on Mobile Agent for Emergent in one transmission. In HGMR, the multicast group is
Event Monitoring divided into subgroups using the mobile geographic
a) Dynamic Route Planning of Mobile Agent: hashing idea: the deployment area is recursively
For an emergent event monitoring scene, when some partitioned into equal-sized square sub-domains called
event occurs, only those nodes in event area would be cells, where d is decomposition index depending on the
activated to cluster. The selection of the next hop for encoding overhead constraints, and each cell consists of
mobile agent not only bases on energy consumption and a manageably-sized subgroup of members. An Access
path loss, but also the stimulated intension received by Point (AP) is responsible for all members in its cell, and
the nodes, in which the discrete emergent event is under APs are managed in turn by a Rendezvous Point (RP).
consideration. The role of each AP or RP is mapped to some unique
b) The Data Aggregation on Mobile Agent geographic location by a simple hash function. The node
The mobile agent consists of identification ID, route that is currently closest to that location then serves the
information, data buffer and processing codes, in which role of AP/RP, and routing to the AP/RP is conveniently
data buffer mainly load the data distilled or fused data achieved by geographic routing. To join a hierarchically
from sensor nodes decomposed multicast group, a node first hashes the
multicast group identifier (GID) to obtain the hashed
IV - Dynamic Minimal Spanning Tree Routing location of the RP via a hashed function and sends a
Protocol (DMSTRP)[4] JOIN message to the RP, which is the same as in the flat
DMSTRP is a cluster-based routing protocol, domain scenario. After receiving the value of the current
uses Minimal Spanning Tree (MSTs) to replace clubs to d of the hierarchy from the RP, the node utilizes the hash
connect the node in the clusters in two layers of the function with d and the node’s location to compute the
network: intra-cluster and inter-cluster. Because clubs are hashed location of the AP belonging to its cell. Note that
less effective than a spanning tree in connecting the computing the hashed location assumes that all nodes
nodes if the network area is larger, DMSTRP is an know the approximate geographic boundaries of the
elegant solution in larger network areas. network. After that the source builds an overly tree, the
(Low Energy Adaptive Clustering Hierarchy)LEACH Source → APs tree, whose the vertices are active APs in
chooses clubs as the basic topology of the network, as a topology graph; and an AP → Members overly tree is
shown in Figure 1 and managing clubs does not need also built from the AP, considering each member as the
multi-hops and thus makes the routing path simple. One vertex. 2 d
step further in (Base Station Controlled Dynamic
3. When a source needs to send data packets, it utilizes the broadcasts data packets to all nodes within its coalition
unicast-based forwarding strategy belonging to HRPM to and looks for the next stage coalition to forward the
propagate data packets to each AP along the Source → packet to. Once the next stage CH, denoted by CHk, was
APs tree. In each cell, adjusting the value of d, the chosen, CHi coordinates the nodes within its coalition to
number of members for which an AP is responsible does cooperatively forward the packet to CHk. This process
not increase too much. Therefore, GMR’s cost over continued until the data were forwarded to the destination
progress optimizing the broadcast algorithm, which is
used to select the next relay node at each hop, contributes
to reduce the number of data transmissions while
maintaining a low encoding overhead compared with the
unicast communication. Sensor nodes running GMR use
the position of their neighbors to select the subgroup
which is the best one to deliver the message towards the
destination, and the selected neighbors can reduce most
the total route to destination. When no neighbor of the
current node can reduce the route to the destination, face
routing is used to circuitously search the path to the
destination. In HGMR, the geographic hashing algorithm
makes the membership management very simple with
almost zero cost. According to the number of the nodes
which play the different roles, HGMR selects the
transmission methods for different hierarchies in reason,
which makes the routing energy-efficient and scalable.
However, the RP is in charge of too much missions in It focus on joint optimal clustering and cooperative
HGMR, which may bring the problem of rapid energy routing. Consider a cooperative sensor network, where a
consumption and make the entire network collapse. node with data would first multicast the packet to a
subset of its neighbors, and then ask them to dynamically
HGMR starts with a hierarchical decomposition of a form a coalition, and cooperatively transmit the packet to
multicast group into subgroups of manageable size (i.e. the next-hop destination. The corresponding energy
encoding overhead) using HRPM’s key concept mobile consumption is the sum of the multicast cost and the
geographic hashing. Within each subgroup, HGMR uses cooperative transmission cost. Intuitively, when the
GMR’s local multicast scheme to forward a data packet number of nodes in a coalition increases, the cooperative
along multiple branches of the multicast tree in one transmission
transmission. Thus, HGMR can simultaneously achieve cost would decrease, but the multicast cost would
energy efficiency (through higher forwarding efficiency increase, and vice versa.
utilizing multicast advantage) and scalability (through
low overhead hierarchical decomposition).
JOINT CLUSTERING AND MINIMUM ENERGY
COOPERATIVE ROUTING includes
VI - Joint Clustering and Optimal Cooperative a) Optimal Coalition Size: Consider a sensor network,
Routing (JCOCR):[6] where each node has a strict power constraint Pmax. Data
need to be routed from a source node S to a destination
joint clustering and optimal cooperative routing, where
node D. In each transmission, an intermediate node
neighboring nodes dynamically form coalitions and
would multicast the packet to a subset of its neighbors,
cooperatively transmit packets to the next hop
and ask the nodes in the subset to dynamically form a
destination. The cooperative sensor network can be
coalition and cooperatively transmit the packet to next
modeled as an edge-weighted graph, based on which
stage destination (point-to-multiple-point transmission
minimum energy cooperative routing is characterized by
first, and then multiple-point-to-point transmission).
using the standard shortest path algorithm.We study two
During the routing process, the number of neighboring
interesting cases: 1) For the case where the delay can be
nodes that participate in the cooperative transmission,
expressed in terms of the number of hops, we use the bi-
i.e., the size of the dynamic coalition, plays a key role.
section method to find the maximum throughput routing;
Note that the energy cost of each transmission is the sum
2) For large scale networks where the end-to-end delay
of the multicast cost and the cooperative cost. Intuitively,
can be approximated as the product of the number of
a larger coalition would reduce the cooperative cost, but
hops and the average one-hop delay, we present a
may require more multicast energy to reach nodes further
polynomial time algorithm to find the maximum
away, whereas a smaller coalition would require less
throughput routing. the energy efficient cooperative
multicast energy but higher cooperative cost. Thus
routing can enhance the performance of WSNs
motivated, we characterize the optimal coalition size to
significantly.
minimize the transmission cost.
We have taken some initial steps to investigate
b) Minimum Energy Cooperative Routing: The minimum
distributed cooperative geographic routing, building on
energy cooperative routing problem (MECR) can be
node cooperation and traditional geographic routing. As
defined as follows. Definition : (MECR) The Instance is
illustrated in Fig. 1, for a given source-destination pair,
given by an edge-weighted directed graph G = (V, E,C, γ)
the routing problem in a coalition-aided network was
and a source destination pair S-D. Let p be a path in G
treated as a multiplestage decision problem, where at
and C(p) be the sum of the costs over the edges on p,
stage i, the coalition head, denoted as CHi, first
C(p) =∑e∈ p C(e). The Problem is to find the optimal
4. path po such that C(po) is minimized. The routing
problem formulated above is a shortest path routing
problem on the new directed graph G, and can be solved
by the well-known Dijkstra’s algorithm. The minimum
energy cooperative routing would achieve better energy
saving, because of the following reasons. 1) It exploits
optimal power allocation within each coalition to reduce
the cooperative transmission cost. 2) It characterizes the
optimal coalition size to minimize the energy cost of
each transmission. 3) It chooses the routing path based
on global information instead of local information.
VII- CONCLUSION
At present, routing in large-scale WSNs is a hot research
topic with a limited but rapidly growing set of efforts
being published. This paper is contribution to study on 4
various routing protocols of Energy-Consumption
Mitigation in large-scale WSN’s.
With the increasing functionalities available to a
wireless sensor node, more complicated tasks which
involve more energy consumption and network overhead
may be assigned to the sensor nodes. To increase energy
efficiency and scalability of the network still remains a
challenging research area.
REFERENCES
[1] Changle Li *, Hanxiao Zhang, Binbin Hao and Jiandong Li,
“A Survey on Routing Protocols for Large-Scale Wireless
Sensor Networks”, www.mdpi.com/journal/sensors
[2] Al-Karaki, J.N.; Kamal, A.E., “ Routing techniques in
wireless sensor networks: A survey”. IEEE Wirel. Commun.
2004, 11, 6-28.
[3] Lingyun Yuan, Xingchao Wang, “Study on Data Gathering
Algorithm Based on MobileAgent and WSN for Emergent
Event Monitoring”, Yunnan Normal University Kunming,
China.
[4] Guangyan Huang1, Xiaowei Li1, and Jing He2, “Dynamic
Minimal Spanning Tree Routing Protocol for
Large Wireless Sensor Networks” Advanced Test Technology
Lab., Institute of Computing Technology,
Chinese Academy of Sciences, Beijing, China, 2006 IEEE
[5] Dimitrios Koutsonikolas1, Saumitra Das1, Y. Charlie Hu1,
and Ivan Stojmenovic2,3, “Hierarchical Geographic Multicast
Routing forWireless Sensor Networks”, The University of
Birmingham, United Kingdom
3SITE, University of Ottawa, Ontario, Canada, 2007 IEEE
[6] Weiyan Ge and Junshan Zhang, Guoliang Xue, “Joint
Clustering and Optimal Cooperative Routing
In wireless sensor network”,IEEE 2008