ENERGY EFFICIENT GRID AND TREE BASED ROUTING PROTOCOL

International Journal of Wireless & Mobile Networks (IJWMN) Vol. 7, No. 3, June 2015
DOI : 10.5121/ijwmn.2015.7302 17
ENERGY EFFICIENT GRID AND TREE BASED
ROUTING PROTOCOL
Reshma J1
, Kavitha C2
and Malashri S3
1
Research Scholar, Department of Computer Science & Engineering,
B N M Institute of Technology, Bengaluru
2
Professor & Head, Department of Information Science & Engineering,
B N M Institute of Technology, Bengaluru
3
M. Tech Student, B N M Institute of Technology, Bengaluru
ABSTRACT
In Wireless Sensor Network, a large number of sensor nodes are deployed and they mainly consume energy
in transmitting data over long distances. Sensor nodes are battery powered and their energy is restricted.
Since the location of the sink is remote, considerable energy would be consumed if each node directly
transmits data to the base station. Aggregating data at the intermediate nodes and transmitting using multi-
hops aids in reducing energy consumption to a great extent. This paper proposes a hybrid protocol
“Energy efficient Grid and Tree based routing protocol” (EGT) in which the sensing area is divided into
grids. The nodes in the grid relay data to the cell leader which aggregates the data and transmits to the
sink using the constructed hop tree. Simulation results show that EGT performs better than LEACH.
KEYWORDS
Wireless Sensor Network, Grid, Cell Leader, Data aggregation, Tree Routing, EGT.
1.INTRODUCTION
Wireless Sensor Network is one of the most important technologies of the twenty first century
[1]. It comprises of a set of sensor nodes and sink nodes and has significant applications in
weather forecasting, military target tracking, medical monitoring and environmental detection
[2][3][4]. It consists of anywhere from a few hundreds to thousands of sensor nodes. However,
these nodes have limited computation capability, energy and memory resource [5]. Since the
sensor nodes are normally battery powered, it is quite challenging to recharge or replace the
batteries.
In Wireless Sensor Network, data transmission takes place in multi-hop fashion where each node
forwards its data to the neighbor node nearer to the sink. In a densely populated network, nearby
sensor nodes sense data that are correlated, thus leading to data redundancy. Data generated is
too much for end-user to process; function for combining data into small set of useful information
is required. In-network filtering and processing techniques can definitely help to conserve the
scarce energy resources [6]. A more practical way of doing this is aggregating data originating
from different nodes in the correlated area. Data aggregation is a process of aggregating sensor
data using data aggregation approaches [7]. A number of aggregation algorithms and database
systems targeting different sensor network scenarios have been proposed [8, 9, 10, 11, 12, 13, 14

18
and 15]. These protocols aim at eliminating redundant data transmission and improving the
lifetime of energy constrained Wireless Sensor Network.
This paper proposes a hybrid protocol called “Energy efficient Grid and Tree based routing
protocol (EGT)”. EGT divides the network into virtual grids. Each grid has a cell leader and is
elected based on the maximum residual energy. The sensor nodes in each grid communicate the
data to its cell leader. The cell leader aggregates the data sent by its members and the aggregated
data is communicated to the sink using Hop Tree. Since this protocol uses both grid and tree
approach, it eliminates transfer of redundant data, thereby minimizing the energy consumption
and maximizing the lifetime of the network. The rest of the paper is organized as follows: In the
next section, we will introduce related work; Section 3 discusses the proposed algorithm. Section
4 presents simulation results and Section 5 presents the conclusions.
2.RELATED WORK
“Low Energy Adaptive Clustering Hierarchy (LEACH)” [16] is one of the most popular
clustering protocols. It forms clusters by using a distributed algorithm. Each node has the same
probability of becoming a cluster-head and the task of being a cluster-head is rotated between
nodes according to a round time. This ensures fair energy dissipation between nodes. A non-
cluster head node in every cluster sends its data to its cluster head. The cluster head compresses
the data received from member nodes and sends the compressed data to the base station.
In LEACH-C [17], clusters are formed by centralized control algorithm and it produces better
clusters by spreading the Cluster Head throughout the network. The steady phase is similar to that
of LEACH, but differs in setup phase. In the setup phase, every sensor node sends its energy
information to remote Base Station and this information is used to select the cluster heads. The
Base Station then broadcasts the ID of cluster head to other member nodes. In this method, the
nodes with more energy have more chances of becoming the cluster head in the current round.
But in this phase, every sensor node needs to send its ID and energy information to remote Base
Station to compete for the role of cluster heads which consumes energy in long distance
transition.
Lindsey et al. proposed “Power-Efficient GAthering in Sensor Information Systems (PEGASIS)”
[18], an energy efficient protocol which provides improvement over LEACH. In PEGASIS, each
node communicates only with a nearby neighbour in order to exchange data. It takes turns in
order to transmit the information to the base station, thus reducing the amount of energy spent per
round. The nodes are organized in such a way so as to form a chain, which can either be formed
by the sensor nodes themselves using a greedy algorithm starting from a certain node or the Base
Station can compute this chain and broadcast it to all the sensor nodes.
Vaibhav V. Deshpande et al. proposed “Energy Efficient Clustering in Wireless Sensor Network
using Cluster of Cluster Heads” [19]. To balance energy consumption among the cluster heads,
this protocol proposes to have cluster of cluster heads within the cluster of sensor nodes. Given a
moment, one cluster head act as master of the given cluster and the master-ship is rotated among
cluster heads after specified number of rounds of communication. This improves the energy
utilization of sensor network, maximizes the network lifetime and makes the Wireless Sensor
Network fault tolerant to some extent.
Neng-Chung Wang [20] proposed “Grid-Based Data Aggregation for Wireless Sensor Networks”
(GBDAS) in which the network is divided into 2-D logical grid of cells. In each cell, the node
with most residual energy is elected as cell head. The cell head from each cell aggregates its own
data with the data sent by all other nodes. All the cell heads are linked to form a chain. The
chain leader is designated based on the most residual energy of all the cell heads. In this protocol,

19
since the cell heads and the chain leader are designated based on the energy level, the energy
depletion of the nodes is distributed evenly.
“An Adaptive Energy aware Data aggregation Tree”(AEDT) [21] uses the maximum energy
available node as the data aggregator. The tree incorporates sleep and awake technology, where
the communicating node and the parent node are in awake state and the remaining nodes go to
sleep state. When the traffic load crosses the threshold value, the packets are accepted adaptively
according to the communication capacity of the parent node. It maintains a memory table which
stores the value of each selected path. Path selection is based on shortest path algorithm where the
node with highest available energy is always selected as the forwarding node.
“A Cluster-based and Tree-based Power Efficient Data Collection and Aggregation Protocol for
Wireless Sensor Networks” (CTPEDCA) [22] is based on clustering and Minimum Spanning
Tree routing strategy for cluster heads. The performance of CTPEDCA is better than that of
LEACH protocol. It prolongs the lifetime of the network and its time complexity is small i.e.,
O(ElogV), where V is the set of cluster heads.
3.PROPOSED ALGORITHM
The proposed system i.e. “Energy efficient Grid and Tree based routing protocol” (EGT)
consists of two phases: Setup Phase and Steady Phase. Setup Phase forms a grid of sensor nodes
and selects a Cell Leader. In each grid, the node with the highest residual energy is elected as the
Cell Leader. In the Steady Phase, the sink node constructs a hop tree algorithm. The sensor
nodes in each of the grid transmit data to the elected cell leaders. These elected cell leaders
aggregate the data and the same is transmitted to the sink using the constructed Hop Tree.
3.1 Cell Leader Election
In the Setup Phase, each node broadcasts its residual energy within the radio range r. Every node
after receiving the message from all its neighbours, updates the neighbourhood table. The node
with the highest residual energy is elected as the cell leader. In addition, each node is randomly
assigned a cost value. When more than one node has the same highest residual energy, then the
node with more cost will be the Cell Leader. Table 1 shows the Cell Leader Election Algorithm.
Table 1. Cell Leader Election
1. Emax  0
2. Nmax NULL
3. Costirand( )
4. for every Ni, iєn(Ni: ID of node)
5. Broadcast available energy Ei within i cluster range
6. Compare Ei and Emax
7. if (Ei > Emax)
8. Emax  Ei
9. Nmax  Ni
10. end if
11. if ( Emax == Ei)
12. if (cost(Ni) >cost(Nmax))
13. Nmax  Ni
14. end if
15. end if
16. end for
17. return Nmax
18. Nmax broadcast status to its neighbors as Cell Leader

20
3.2 Routing Tree Generation
Once the cell leader is elected and the nodes in the cell want to transfer data to the sink, it
becomes essential for the cell leader to communicate the data to the sink using shortest path. A
Hop Tree is constructed for finding the shortest path from the cell leaders to the sink. Table 2
below depicts the Hop Tree Construction Algorithm.
Table 2. Hop Tree Construction Algortihm [23]
1. Every Cell Leader sends its status to Sink
2. HTSsink  0
3. HTSNi  ∞
4. Sink floods HCM( Type, ID,HTS) to CellLeaderList
5. if ( HTS(u) > HTS(HCM)+1)
6. NH(u) = ID(HCM)
7. HTS(u)=HTS(HCM)+1
8. ID(HCM) = ID (u)
9. HTS(HCM)= HTS(u)
10. end if
11. u transmits HCM to its neighbors
The distance from the sink to each node is computed in hops. The sink floods the Hop
Configuration Message (HCM) to all the cell leaders. The HCM message contains three fields:
Type, ID and HopToSink where Type is type of the message i.e., HCM, ID is node identifier that
started or retransmitted the HCM message and HopToSink is the distance in hops by which an
HCM message has passed.
The HopToSink value starts with value zero at the sink and forwards it to its neighbors (at the
beginning, all Cell Leaders set the HopToSink as infinity). Each cell leader upon receiving the
HCM message verifies if the HopToSink value in the HCM message is less than its own. If so,
the node updates its NextHop value with the value of the ID field in HCM message. In addition,
the values of HopToSink variable and ID field of HCM message is updated as in Table 2. Cell
leader now relays the updated HCM message. Otherwise, the cell leader discards the received
HCM message. The above steps are repeated until the whole network is configured. All the
updated information i.e. NextHop and HopToSink for every Cell Leader is stored in routing table
rTable.
3.3 Intra-Cluster Communication
In their allotted time slot, the non-leader nodes in each grid communicate the sensed data to their
respective cell leader. Once all the cell members in the network finish transmitting their sensed
data, inter-cluster communication phase begins. In a densely populated sensor network, the area
detected by sensors may overlap and thus the data sensed may be correlated. This causes data
redundancy at the cell leader during intra-cluster communication. Moreover, much energy is
consumed in processing similar data and results in more energy consumption. Hence, data
aggregation is performed during intra-cluster communication.
Table 3 shows the Intra-Cluster Communication Algorithm. The nodes in each grid transmit the
sensed temperature data to the elected Cell Leader. The Cell Leader in each grid aggregates its
own data with the data received from all other nodes in that cell using an average function.

21
Table 3. Intra-Cluster Communication Algorithm
1. n  no. of Cell Member in each cell
2. for every cell Ci
3. for every Cell Member j in Ci
4. send temperature( CM j ) to Ch
5. end for
6. Sum  ∑j=1 to n temperature( CM j )
7. Avg( CL I )  Sum/n
8. end for
Figure 1 below depicts Intra-Cluster Communication. Each node is labeled with node id followed
by its residual energy level. Taking the example of bottom rightmost grid, it can be observed that
the node with id 42 has the highest residual energy of 95 Joules and hence it is elected as cluster
head. All the other nodes in that grid send its temperature data to the node with id 42. This node
aggregates its data with the data of all other nodes using average function.
C1 C4
C2 C3
C5 C6
Figure 1. Intra-cluster Communication
3.4 Inter-cluster Communication
In Inter-Cluster Communication, routing take place between cell leaders of different cells and the
data is relayed to the sink. Paths are established from every cell leader to sink using Hop Tree
Construction(HTC) Algorithm of Table 2 and the next hop information for every cell leader is
updated in the routing table. Every cell leader accesses the next hop information and forwards its
data until its HopToSink (HTS) count decreases to zero. Table 4 below depicts the Inter-Cluster
Communication Algorithm.
Table 4. Inter-cluster Communication Algorithm
1. for every CLi in routing table rTable
2. src  CLi
3. while (hoptosink(CLi ≠ 0))
4. send Avg(CLi) from src to nexthop(src)
5. hoptosink(CLi)= hoptosink(CLi) - 1
6. src  nexthop(src)
7. end while
8. end for
11,96J
12,95.5J
13,96.5J
14,97.5J
10,97J
8,97.5J
30,94J 33,95J
29,96J
31,95J
30,96.5J
32,95.5J
35,97J
37,93J
40,92.5J
41,92J
42,95J
36,94J
39,93J38,94.5J
28,95.5J
25,95J
22,94J
24,96J
26,96.5J
27,97J
23,94.5J
2,97.5J
7,99J
5,98J
3,98.5J6,98.5J
1,97.5J
4,97J
18,97.5J
9
5
h
g
g
20,96J
17,98J
16,96.5J
19,97J
15,95J
21,97.8J
Cluster head
Sink
Intra cluster
communication
9,98J
43,200J
Cluster member

22
Figure 2 below depicts how the data aggregated from each of the Cell Leader is sent to the sink
using the constructed Hop Tree. The aggregated data from Cell Leader C4 and C6 is sent to Cell
Leader C3. The Cell Leader C3 aggregates its own data with that of C4 and C6 and sends it to
Cell Leader C1. The Cell Leader C1 aggregates its own data with that sent by C3 and the same is
dispatched to the sink.
C1 C4
C2 C3
C5 C6
Figure 2. Inter-cluster communication using Hop Tree constructed
4. SIMULATION RESULTS
The simulation environment models a grid based network in which 50 nodes are deployed
randomly or uniformly over an area of 1500 x 1500 m2
. The simulation is carried out using NS2.
NS2 is an open source event driven simulator tool. Table 5 below shows the list of parameters
used for simulation.
Table 5. Simulation Settings
Simulation
Parameters
Values
Topology size 1500 x 1500 m2
Number of sensors 50
Deployment type Random / uniform
Transmission range 500 m
Data Packet Size 1000 bytes
Traffic Type Constant Bit Rate
MAC protocol 802.15.4
Initial energy 100 J
11,96J
12,95.5J
13,96.5J
14,97.5J
10,97J
8,97.5J
30,94J 33,95J
29,96J
31,95J
30,96.5J
32,95.5J
35,97J 37,93J
40,92.5J
41,92J
42,95J
36,94J
39,93J
38,94.5J
28,95.5J
25,95J
22,94J 24,96J
26,96.5J
27,97J
23,94.5J2,97.5J
7,99J
5,98J
3,98.5J6,98.5J
1,97.5J
4,97J
18,97.5J
9
5
h
g
g
20,96J
17,98J
16,96.5J
19,97J
15,95J
21,97.8J
9,98J
43,200J

23
Simulation results show the comparison of EGT with LEACH protocol. Figure 3 shows the
performance of throughput varying with simulation time. The graph clearly shows that EGT has a
greater throughput compared to LEACH. LEACH assumes direct communication between
cluster head and sink. Data destined from distant cluster heads gets dropped and hence throughput
is less. Whereas in EGT, the cell leaders use multi-hop transmission to relay data to the sink and
this results in higher throughput.
Figure 3. Throughput comparison
Figure 4 shows the performance of EGT compared to LEACH based on packet delivery ratio. At
time nine seconds, the cluster heads aggregate the data from the respective grid and transmit to
the sink. EGT uses hop tree for inter-cluster communication and hence the aggregated data
transmitted by every cluster head will be successfully received by the sink. Therefore, it can be
observed from the graph that the packet delivery ratio is 1 for EGT, meaning that the number of
packets generated in the network is successfully received at the destination. On the other hand,
LEACH assumes direct communication between every cluster head and the sink. As a result,
packets transmitted by cluster heads which is out of the coverage area of the sink will be dropped
and hence its packet delivery ratio is less than 1. Therefore, the proposed protocol has a greater
packet delivery ratio.
Figure 4. Delivery Ratio Comparison
EGT
LEACH
LEACH
EGT

24
Figure 5 shows the average energy remaining in the network after data transmission. In LEACH,
since the cluster heads are elected using probabilistic approach, there is a possibility that nodes
with lesser remaining energy may be chosen and may die first. As a result, data may have to be
retransmitted resulting in more energy consumption. However, such a situation does not arise
with EGT and hence EGT has more average energy remaining in the network, after the data is
transmitted to the sink.
Figure 5. Average Energy Comparison
5. CONCLUSION
In a Wireless Sensor Network, the sensor nodes are battery powered and this act as a major
constraint on the energy. Energy is consumed in sensing and transmission. In a densely
populated network, nearby nodes sense the same information and this results in transmission of
redundant data to the sink. To conserve energy, it is essential to employ energy efficient routing
techniques and to avoid transmitting redundant data to the sink. This paper proposes an energy
efficient hybrid protocol named “Energy efficient Grid and Tree based routing protocol” (EGT).
EGT divides the sensing area into grids. The cell leader from each grid receives the sensed data,
aggregates it and transmits it to the sink using the constructed hop tree. Simulation results show
that the proposed protocol is more energy efficient compared to LEACH protocol.
REFERENCES
[1] “21 ideas for the 21st Century”, Business Week, August 30 1999, pp 78-167.
[2] I.F. Akyildiz and M.C. Vuran, Wireless Sensor Networks, Advanced Texts In Communications And
Networking, p.480, 2010.
[3] D. Pompili,T Melodia and I.F. Akyildiz, Three-dimensional and two-dimensional deployment
analysis for underwater acoustic sensor networks, Ad Hoc Networks, 7(4):778-790,2009.
[4] L. Li and T. Kunz, Cooperative Node Localization for Tactical Wireless Sensor Networks, Proceed-
ings of IEEE MILCOM'07, 2007.
[5] I.Gupta, D.Riordan and S.Sampalli,“Cluster–head Election using Fuzzy Logic for Wireless Sensor
Networks”,Communication Networks and Services Research Conference, pp.255-260, May 2005.
[6] Manju Bala and Lalit Awasthi,“Proficient D-HEED Protocol for Maximizing the Lifetime of WSN
and Comparative Performance Investigations with Various Deployment Strategies”, International
Journal of Advanced Science and TechnologyVol.45,August, 2012.
LEACH
EGT

25
[7] Nandini. S. Patil and Prof. P. R. Patil, “Data Aggregation in Wireless Sensor Network”, IEEE
International Conference on Computational Intelligence and Computing Research,Coimbatore, Dec
2010.
[8] S. Madden,R. Szewczyk,M. FranklinandD. Culler,“Supporting aggregate queries overad-hoc
sensornetworks”, In Workshopon Mobile Computingand Systems Applications (WMCSA),Callicoon,
NY, June 2002, Pages: 49 – 58.
[9] Y. Yao and J. Gehrke, “Query Processing in Sensor Networks”, In Proceedings of the First Biennial
Conference on Innovative Data Systems Research (CIDR 2003),Asilomar,California,January2003.
[10] Demers, J. Gehrke, R. Rajaraman, N. Trigoni, and Y. Yao, “Energy-Efficient Data Management for
Sensor Networks: A Work-In-Progress Report”, 2nd IEEE Upstate New York Workshop on Sensor
Networks. Syracuse, NY, October 2003.
[11] S. Madden, M. Franklin, J. Hellerstein, W. Hong, “The Design of an Acquisitional Query Processor
for Sensor Networks”, ACM SIGMOD Conference, San Diego, CA, June 2003, Pages: 491 - 502.
[12] Y. YaoandJ. Gehrke, “The Cougar Approach to In-Network Query Processingin Sensor
Networks”,Sigmod Record, Volume 31, Number 3, September 2002, Pages: 9 – 18.
[13] J. Heidemann,F. Silva,C. Intanagonwiwat,R. Govindan,D. EstrinandD. Ganesan,“Building efficient
wireless sensor networks with low-level naming”, In Proceedings of the eighteenth ACM
symposiumon Operating systems principles,Banff, Alberta,Canada, October 2001, Pages: 146-159.
[14] C. Intanagonwiwat, D. Estrin, R. Govindan, and J. Heidemann, “Impact of network density on data
aggregation in wireless sensor networks”, in ICDCS-22, November 2001.
[15] C. Intanagonwiwat, R. Govindan andD. Estrin,“Directed diffusion:A scalableand robust
communication paradigm for sensor networks”,In Proceedings of the Sixth Annual International
Conferenceon Mobile Computing and Networks (MobiCOM 2000),Boston,MA,August 2000, pages:
56-67.
[16] W R Heinzelman, A Chandrakasan and H. Balakrishnan, “Energy-efficient communication protocol
For wireless microsensor networks,”in Proceedings of the33rdAnnualHawaii International
Conference on System Sciences (HICSS '00), January 2000.
[17] W. B. Heinzelman, A. P. Chandrakasanand H. Balakrishnan, “An application-specific protocol
architecture for wireless microsensor networks,”IEEE Transactions on Wireless Communications,
vol. 1, no. 4, pp. 660–670, 2002.
[18] Lindsey S and Raghavendra C., “PEGASIS: Power-Efficient GAthering in Sensor information
Systems”, In Proceedings of the IEEE Aerospace Conference, Los Angeles, MT, USA, 2002; pp.
1125–1130.
[19] Vaibhav V. Deshpande,Arvind R. Bhagat Patil,“Energy Efficient Clusteringin Wireless Sensor
Network using Cluster of Cluster Heads”, 10thInternationalConference on Wireless and Optical
Communications Networks (WOCN) , 26-28th July 2013, Bhopal, pp 1-5.
[20] Neng-Chung Wang et. al, “Grid-Based Data Aggregation for Wireless Sensor Networks” (GBDAS),
Journal of Advances in Computer Networks, Vol 1, No. 4, December 2013.
[21] Deepali Virmani,TanuSharma,“An Adaptive Energy aware Data aggregation tree” (AEDT),Science
and Development Network (SciDev.Net), Jan 2013.
[22] Wei Wang, Bingwen Wang Zuo Liu, Lejiang Guo and Wei Xiong, “A Cluster-based and Tree- based
Power Efficient Data Collection and Aggregation Protocol for Wireless Sensor
Networks”,Information Technology Journal, Jan 2011; 10: 557-564.
[23] Yalin Nie, Sanyang Liu, Zhibin Chen and Xiaogang Qi, “A Dynamic Routing Algorithm for Data-
Aggregation Optimization in event-Driven Wireless Sensor Networks”, Journal of Communications
Vol. 8, No. 8, August 2013.
AUTHORS
Reshma J received B.E and M Tech degree in Computer Science and Engineering from
Bangalore University and Visvesvaraya Technology University respectively. She has more
than ten years of teaching experience. Her research interests include wireless networks and
sensor networks.

26
Dr.Kavitha C received her B.E. and M.E degree in Computer Science & Engineering from
Bangalore University, Karnataka India and her Ph.D from Visvesvaraya Technological
University, Belgaum, Karnataka – India in 2010. She has more than eighteen years of
teaching experience. Her research interests include wireless networks, sensor networks.
Malashri S received B.E degree in Information Science and Engineering from Visvesvaraya
Technology University in 2011 and pursuing M. Tech degree in Computer Network
Engineering in Visvesvaraya Technology University. She has two years of teaching
experience. Her research interests include Wireless networks and sensor networks.

ENERGY EFFICIENT GRID AND TREE BASED ROUTING PROTOCOL

More Related Content

What's hot (17)

Viewers also liked (17)

Similar to ENERGY EFFICIENT GRID AND TREE BASED ROUTING PROTOCOL (20)

Recently uploaded (20)

ENERGY EFFICIENT GRID AND TREE BASED ROUTING PROTOCOL