E-HORM: An Energy-efficient Hole Removing Mechanism in Wireless Senor Networks

Abstract

Cluster based routing protocols forWireless Sensor Networks (WSNs) have
been widely used for better performance in terms of energy efficiency.
Efficient use of energy is challenging task of designing these
protocols. Energy holes are created due to quickly drain the energy of a
few nodes due to nonuniform node distribution in the network. Normally,
energy holes make the data routing failure when nodes transmit data back
to the sink. We propose Energy-efficientHOle Removing Mechanism (E-HORM)
technique to remove energy holes. In this technique, we use sleep and
awake mechanism for sensor nodes to save energy. This approach finds the
maximum distance nodes to calculate the maximum energy for data
transmission. We consider it as a threshold energy Eth. Every node first
checks its energy level for data transmission. If the energy level of
node is less than Eth, it cannot transmit data.

Full text

E-HORM: AN ENERGY-EFFICIENT HOLE REMOVING MECHANISM
IN WIRELESS SENSOR NETWORKS
‡ ‡ $♯§
‡COMSATS Institute of Information Technology, Islamabad, Pakistan.
$Faculty of Engineering, Dalhousie University, Halifax, Canada.
♯University of Alberta, Alberta, Canada.
§King Abdulaziz University, Rabigh, Saudi Arabia.
ABSTRACT
Cluster based routing protocols for Wireless Sensor Networks
(WSNs) have been widely used for better performance in terms
of energy efficiency. Efficient use of energy is challenging
task of designing these protocols. Energy holes are created
due to quickly drain the energy of a few nodes due to non-
uniform node distribution in the network. Normally, energy
holes make the data routing failure when nodes transmit data
back to the sink. We propo se Energy-efficient HOle Removing
Mechanism (E-HORM) technique to remove energy holes. In
this technique, we use sleep and awake mechanism for sensor
nodes to save energy. This approach finds the maximum dis-
tance nodes to calculate the maximum energy for data trans-
mission. We consider it as a threshold energy Eth. Every
node first checks its energy level for data transmission. If the
energy level of node is less than Eth, it cannot transmit data.
—Energy Hole; Non-uniform Distribution;
Corona; WSNs
1. BACKGROUND
WSNs consist of a large number of sensor nodes deployed
in a sensor field for object monitoring either inside the field
or near the field. Recent advances in Microelectromechan-
ical Systems (MEMS) based technologies have enabled the
deployment of a large number of tiny sensor nodes with lim-
ited battery life time. These nodes have the low computa-
tional ability and small internal memory. These small sensor
nodes capable of monitoring, sensing, aggregation and trans-
mission of data to the sink. WSNs are used in many commu-
nication applications including security, medical, surveillance
and weather monitoring. Sensor nodes are able to measure
various parameters of the environment and transmit collected
data to the sink directly or through multihop communication.
Nodes deployment is the first step in establishing sensor net-
work. Sensor nodes are battery powered and randomly de-
ployed in target area. After the deployment sensor network
cannot perform manually. Optimizing the energy consump-
tion is one of the major tasks in WSNs to prolong the network
lifetime. To address this issue, much work has been done in
this area during the last few years. If the sensor nodes are de-
ployed uniformly, nodes near the sink send their own data as
well as the date collected by other nodes away from the sink
in multihop scenario. In this case, the sensor nodes near the
sink consume more energy and die out quickly. As a result,
the sensor network will disconnect having sufficient energy
left unused [1]. Various schemes have been proposed to ad-
dress the Energy Hole Problem (EHP). In [1], authors present
a model for balanced density control to avoid energy holes.
They use equivalent sensing radius and pixel based transmis-
sion schemes for balanced energy consumption. By activating
different energy layered nodes in non-uniform distribution the
energy holes problem is mitigated effectively.
In [2], authors purpose a Voronoi diagram-based distri-
bution model for sensor deployment. In this model, each
node calculate its Voronoi polygon to detect coverage hole
and move towards a better position for maximum coverage in
the field.
In [3], authors discuss the Corona-based sensor network
model for balanced energy depletion due to many to one com-
munication in multihop sensor networks. They use mobile
sensors to heal the coverage hole created due to large data re-
laying near the sink. In [4], authors purpose a multiple sink
model to divide the network load near the sink to avoid the en-
ergy hole. Decision of a multiple sink is based on the amount
of data loads in sensor networks.
In [5], authors discuss the distributed localization prob-
lem, Optimal Geographic Density Control (OGDC) for full
coverage as well as connectivity. They prove that if commu-
nication range is twice the sensing range then complete cov-
erage implies connectivity.
Tang and Xu [2], discuss how to optimize the network
lifetime and data collection at the same time. Large amount
of data is given to the sink by nearby nodes and less data from
the nodes that are far away from the sink in the previous study.
For this a rate allocation algorithm Lexicographical Max-Min
(LMM) for data gathering is proposed to maximize the data-
gathering amount and maximize the network lifetime under
2013 26th IEEE Canadian Conference Of Electrical And Computer Engineering (CCECE)
978-1-4799-0033-6/13/$31.00 ©2013 IEEE

balance data gathering condition.
2. ENERGY HOLE PROBLEM IN WSNS
In this section, we discuss the characteristics and effects of
energy holes. Unbalanced energy consumption is a major is-
sue in WSNs when nodes are randomly deployed in sensor
networks. Sensor nodes in the network behave as a data orig-
inator and data router [6].
Nodes near the sink have a greater load of data, hence
consume more energy. Therefore, nodes near the sink de-
plete more energy and die quickly, leading what is called EHP
around the sink. In this situation, no more data will be trans-
mitted to the sink. So, the network lifetime ends due to more
depletion of energy near the sink. More sensor nodes due to
dense deployment in any region may overlap and increase the
hardware cost. However dense deployment is another reason
for the creation of holes problem in WSNs.
3. ENERGY CONSUMPTION MODEL
We have found that due to EHP network dies early.
[7] deals with the lifetime of sensor networks. They as-
sume that the communication between nodes consume more
energy rather than data aggregation and data reception. Pre-
vious research shows that the traffic near the sink is heavier
than the traffic away from the sink. Energy consumption near
the sink is greater and energy holes occur leading to the death
of the entire network. This phenomenon reduces the lifetime
of the whole network due to large energy consumption near
the sink. If more nodes are deployed near the sink, there will
be more nodes use to relay the distant data and hence extends
the network lifetime which is also non-uniform node distribu-
tion strategy. The phenomenon of an energy hole makes the
researchers realize that the network lifetime is determined by
the weakest node.How to avoid EHP becomes an important
research area now a days.
We use the same energy consumption model as used in [8],
which is the first-order radio model.
4. E-HORM: PROPOSED SCHEME
We consider the scenario where nodes are randomly deployed
in a given region. Some nodes are selected to be active and
rest are in sleep mode to maintain sensing, coverage and con-
nectivity.The position of the sink is at the center of the net-
work. Energy of all the nodes are equal while the energy of
the sink is unlimited. In our model, nodes transmit data to
the sink based on residual energy and the distance between
nodes and sink. E-HORM scheme has four major phases: (i)
initializing phase, (ii) threshold calculating phase, (iii) clus-
ter formation, and (iv) sleep/awake scheduling phase. In ev-
ery round sink first checks the maximum distance node in the
field. It then calculates the required energy to transmit data to
the sink. We set this energy as a Eth. In every round if the
energy level of a node is greater than or equal to Eth , sensor
node transmits data to the sink. If the energy level of any node
is less than Eth, it cannot transmit data to the sink. When the
energy level is less than Eth value, it moves towards the sleep
mode to save energy.
4.1. Sensor Node Sleep Scheduling
Before performing the sleep schedule, we examine the energy
level of each node according to their distance from sink by
using the following steps.
•Case 1: Er> Eth : When Eris greater than the Eth ,
the node is in active mode and ready for communica-
tion.
•Case 2: Er< Eth : When Eris less than the Eth , the
node moves towards sleep position.
Each node set the sleeping scheduling according to the Eth.
To calculate the Eth , we use the following formula.
Eth = ((ET X +EDA)∗D) + (Eamp ∗D∗d4)(1)
Where Dis the length of data packet and dis the distance
between maximum distance node and sink.
4.2. E-HORM Formulation
Based on the network model, nodes belonging to CH forward
both the data generated by themselves and the data generated
by its member nodes. Nodes which are not CH need not for-
ward any data. Suppose nodes are randomly deployed in the
network and there is no need for data aggregation at any for-
warding node. Based on transmission mechanism data for
CH receive and forward is (D1+D2+D3+.... +DN)and
(DCH +D1+D2+D3+.... +DN). If the distance be-
tween N and CH is d < d0than energy consumption for data
transmission from N to the CH.
ECH
N=DCH
N(Eele) + DC H
N(Efs )(d2)(2)
Now considering the scenario where distance between N to
CH is d > d0.
ECH
N=DCH
N(Eele) + DC H
N(Eamp)(d4)(3)
Energy consumed by CH to transmit data to the S when dis-
tance between them is d < d0:
ES
CH =DS
CH (Eele ) + EDA +DS
CH (Ef s )(d2)(4)
When distance between CH and S is d > d0:
ES
CH =DS
CH (Eele ) + EDA DCHS(Eamp )(d4)(5)

0 500 1000 1500 2000
0
20
40
60
80
100
No. of rounds (r)
No. of nodes dead
iLEACH
LEACH
0 500 1000 1500 2000
0
2000
4000
6000
8000
10000
12000
14000
No. of rounds (r)
No. of Pkts to BS
iLEACH
LEACH
Fig. 1. Comparing the performance of LEACH and iLEACH
0 500 1000 1500 2000 2500 3000 3500
0
20
40
60
80
100
No. of rounds (r)
No. of nodes dead
iTEEN(hard)
TEEN(hard)
0 1000 2000 3000 4000 5000
0
2000
4000
6000
8000
10000
12000
14000
No. of rounds (r)
No. of Pkts to BS
iTEEN(hard)
TEEN(hard)
Fig. 2. Comparing the performance of TEEN and iTEEN
ET otal−CH =EC H +EN(6)
EAverage−C H =ET otal−C H
N(7)
Energy saving in each round for normal node is:
ESave−N=Eelec +ET X +Eamp (8)
Energy saving for CH is:
ESave−C H =Eele +EDA +ET X +ERX +Eamp (9)
Energy saving for all sleep nodes
ESave−T otal =
n
X
i=0
Ei(10)
4.3. Analysis
Sleep probability of sensor nodes for each round is based on
Eth of distant nodes. The nodes away from the sink increase
the sleep probability. In this way, the nodes consume approx-
imately balance energy to enhance the network lifetime. For
WSNs the sleeping scheduling is very important due to lim-
ited energy of sensor nodes. If a node set into active position
for a long time, it consumes a lot of energy. On the other way,
the transmissions create more delay for long time sleep du-
ration. In this paper, we design an optimum sleeping control
mechanism to avoid both of the problems.
5. SIMULATION RESULTS
Our results are based on analyses and are validated by Matlab
simulations. In our simulations, we consider a sensor net-
work with nnumber of homogeneous or heterogeneous sen-
sor nodes which are randomly deployed in a square field. The
0 500 1000 1500 2000
0
20
40
60
80
100
No. of rounds (r)
No. of nodes dead
iSEP
SEP
0 500 1000 1500 2000
0
5000
10000
15000
No. of rounds (r)
No. of Pkts to BS
iSEP
SEP
Fig. 3. Comparing the performance of SEP and iSEP
0 500 1000 1500 2000 2500 3000 3500
0
20
40
60
80
100
No. of rounds (r)
No. of nodes dead
iDEEC
DEEC
0 500 1000 1500 2000 2500 3000 3500
0
1
2
3
4
5
6
7
8
9x 104
No. of rounds (r)
No. of Pkts to BS
iDEEC
DEEC
Fig. 4. Comparing the performance of DEEC and iDEEC
only sink is located at the center of the field. Sensor nodes
do not move after deployment. Sensor nodes are limited ini-
tial energy Einit , while the energy of the sink is unlimited.
The transmission ranges of sensor nodes are adjustable ac-
cording to the distance from the sink. All nodes need to send
the data packets to the sink in a cycle time. In each round sen-
sor nodes are selected to work and the rests of nodes are set
to sleep mode to save energy. In E-HORM, this mechanism is
called sleep awake process. We apply our scheme in four cat-
egories of the cluster based protocols LEACH, TEEN, DEEC
and SEP.
Table 1.Simulation Parameters
Symbol Description Value
XmDistance at x-axes 100 meter
YmDistance at y-axes 100 meter
N Total number of nodes 100 Nodes
E0Total energy of node 0.5 j
PProbability of cluster head 0.1
ERX Energy dissipation: receiving 0.0013/pj/bit/m4
Efs Energy dissipation: free space model 10/pj/bit/m2
Eamp Energy dissipation: power amplifier 100/pj /bit/m2
Eele Energy dissipation: electronics 50nj/bit
ET X Energy dissipation: transmission 50/nj/ bit
EDA Energy dissipation: aggregation 5/nj/bit
d0Reference distance 87 meter
nNumber of sleep nodes 10 Nodes
ErRemaining energy of nodes
In this section, we evaluate and explain the simulation re-
sults of our purposed work. Three main matrices, including
the stability period, network survival lifetime and data trans-
mitted to the sink are calculated and compared with the ex-
isting work. We explore the theoretical analysis of node dis-
tribution according to random deployment. We evaluate and
compare our purposed technique with LEACH, TEEN, DEEC
and SEP protocol with random node deployment in the sensor

network. Our technique outperforms in terms of network life
time and stability period. Now we explain how the sleep and
awake mechanism are carried out in LEACH and TEEN to re-
move the energy holes. LEACH and TEEN are homogeneous
routing protocol and all the nodes have the same probability
to become a cluster head.
Cluster head consumes more energy during transmission due
to data load of member nodes. Nodes forward their own data
to cluster heads according to TDMA schedule. All nodes are
in sleep node and turn on their transmitters during data trans-
mission to save energy. According to our approach, the nodes
that have the energy level less than the threshold are in sleep
mode to save energy. In this way we save energy to prolong
network life time and stability period.
SEP and DEEC are the heterogeneous clustering proto-
col. In this section, we compare the performance of SEP
and DEEC with our proposed scheme iSEP and iDEEC in
the same heterogeneous setting. Extra energy is distributed
over-all advance nodes in the field. This setting latter provide
the more stable region and network lifetime. Fig. 3 shows the
result of SEP and our proposed scheme iSEP. It is very clear
that the stable region of iSEP is greater than SEP even though
the network lifetime is not very large. Due to balanced en-
ergy consumption network, stability increased. In SEP, when
first node dies the system becomes unstable due to population
reduction. The death ratio of normal nodes are greater than
advance nodes due to having more energy of advance nodes.
Distant nodes consume more energy and are put into sleep
mode to avoid an energy holes. In each round, the nodes hav-
ing less energy then Eth put into sleep mode to save energy.
This mechanism enhances the stability period of SEP due to
better utilization of energy. Now we evaluate and compare
the performance, of DEEC and iDEEC protocols. For best
evaluation of performance, we ignore the disturbances due
to signal collision and interference in a wireless medium. It
is obvious that the network lifetime of our proposed scheme
is grater compared with DEEC protocols. There is a minute
change in the stability period of DEEC and iDEEC while, the
network lifetime of iDEEC is greater. This is because the
energy of each node is different from other nodes. In peri-
odically sleep and awake mechanism, we can best utilize the
energy consumption in each region to remove energy holes.
6. CONCLUSION
In this article, we focus on energy hole problem and energy
consumption in LEACH, TEEN, DEEC and SEP protocols.
We discussed the creation of energy holes in homogeneous
and heterogeneous routing protocols. We implement our ap-
proach in LEACH, TEEN, DEEC and SEP routing protocol.
Due to random deployment in these protocols, there exists the
probability of energy holes. Sleep and awake mechanism to
remove energy holes in WSNs is proposed. We investigated
that after our proposed scheme, a better energy consumption
is achieved. As for network life time, this work clearly gives
the results in terms of network lifetime and stability period.
Sensor nodes consume balance energy, and hence maximize
the network lifetime. This paper clearly points out how we
can remove the energy holes problem in WSNs, and other re-
searchers can also easily purpose a new protocol according to
deployment techniques, which avoid the energy holes. Sim-
ulation results show that the results of our scheme perform
better than previous schemes in terms of network life time
and stability perid.
7. REFERENCES
[1] J. Jia, J. Chen, X. Wang, and L. Zhao, Energy-balanced
density control to avoid energy hole for wireless sen-
sor networks, International Journal of Distributed Sen-
sor Networks, vol. 2012, 2012.
[2] X. Tang and J. Xu, Optimizing lifetime for continuous
data aggregation with precision guarantees in wireless
sensor networks, IEEE/ACM Transactions on Network-
ing (TON), vol. 16, no. 4, pp. 904917, 2008.
[3] J. Jia, X. Wu, J. Chen, and X. Wang, Exploiting sensor
redistribution for eliminating the energy hole problem in
mobile sensor networks, EURASIP Journal on Wireless
Communications and Networking, vol. 2012, no. 1, p.
68, 2012.
[4] M. Ahadi and A. Bidgoli, A multiple-sink model for de-
creasing the energy hole problem in large-scale wireless
sensor networks,
[5] H. Zhang and J. Hou, Maintaining sensing coverage and
connectivity in large sensor networks, Ad Hoc & Sensor
Wireless Networks, vol. 1, no. 1-2, pp. 89124, 2005.
[6] I. Akyildiz, W. Su, Y. Sankarasubramaniam, and E.
Cayirci, A survey on sensor networks, Communications
Magazine, IEEE, vol. 40, no. 8, pp. 102114, 2002.
[7] G. Chen, C. Li, M. Ye, and J. Wu, An unequal clus-
terbased routing protocol in wireless sensor networks,
Wireless Networks, vol. 15, no. 2, pp. 193207, 2009.
[8] J. Li and P. Mohapatra, An analytical model for the
energy hole problem in many-to-one sensor networks,
in IEEE Vehicular Technology Conference, vol. 62, p.
2721, IEEE; 1999, 2005.