International Journal of Computer Science & Information Technology (IJCSIT) Vol 4, No 1, Feb 2012

A CLUSTER ALLOCATION AND ROUTING

ALGORITHM BASED ON NODE DENSITY FOR
EXTENDING THE LIFETIME OF WIRELESS SENSOR
NETWORKS
Wernhuar Tarng, Hao-Wei Lin, and Kuo-Liang Ou

Graduate Institute of Computer Science, National Hsinchu University of Education,

Hsinchu, Taiwan
wtarng@mail.nhcue.edu.tw, hill5201908@gmail.com, klou@mail.nhcue.edu.tw

ABSTRACT
The electricity of sensor nodes in wireless sensor networks is very limited, so it is an important research
topic to deploy the sensor nodes and cooperate with an efficient routing algorithm for extending the
network lifetime. In the related research, LEACH routing algorithm randomly selects cluster heads in
each round to form a cluster network, which may cause additional power consumption and inability to
maintain the optimal routes for data transmission. The cluster allocation and routing algorithm proposed
in this study is based on the cluster architecture of LEACH, and the objective is to produce clusters with
more sensor nodes to balance energy consumption of cluster head. For indirect-transmission routing
algorithms, the sensor nodes near the base station may consume more energy due to a larger amount of
data transmission. Therefore, this study proposed to increase the node density near the base station
during deployment to compensate for the requirement of high energy consumption. The experimental
results show that the proposed algorithm based on node density distribution can efficiently increase the
lifetime of wireless sensor networks.

KEYWORDS
Wireless Sensor Networks, Routing Algorithms, Cluster Networks, Node Density, Network Lifetime

1. INTRODUCTION
The wireless sensor network (WSN) was initially a research project directed by UC Berkeley,
where they used the micro-electro-mechanical technology to design wireless sensors about the
size of a coin, also called smart dusts. A WSN consists of spatially distributed autonomous
sensors to monitor physical or environmental conditions, such as temperature, sound, vibration,
pressure, pollutants and motion to cooperatively pass their data through the network to the base
station [1]. The WSN was first used to collect data in the battle fields. Due to its small size, low
price and various functions, the WSN has now been used in daily applications, for example,
taking care of elder people, vibration detection of bridge piers, detection of fire and CO2
emission, and so on.

In the data transmission of WSN, the strength of signals is highly related to the distance, i.e., the
longer the distance, the weaker the signal and the more energy consumed. When the distance is
too long, it may cause the sensor nodes to exhaust energy quickly and thus affect the network
lifetime. As a result, how to fully utilize the electric power of sensor nodes for continuous data
transmission is an important topic of WSN research. Basically, the ways of data transmission
for the WSN can be categorized as direct transmission and indirect transmission [2].

DOI : 10.5121/ijcsit.2012.4104 51
 International Journal of Computer Science & Information Technology (IJCSIT) Vol 4, No 1, Feb 2012
A. Direct Transmission
Each sensor nodes transmit the collected data to the base station directly. In this approach, the
data rate is higher since there is no forwarding during transmission. Although the
implementation for direction transmission is easier, the data may not be sent to the destination
when the application area is very large, or the sensor nodes may die quickly due to high energy
consumption. Therefore, this approach is suitable only for small-area applications.
B. Indirect Transmission
The sensor nodes send the collected data to the neighbour nodes, which will forward the data to
the base station. This approach solves the problem of high energy consumption in long-distance
transmission, but it may cause some nodes to consume more energy by forwarding data for the
other nodes and thus affecting the network lifetime.
Usually, the deployment of a WSN is to scatter a large amount of sensor nodes over the
application area randomly. Then, the sensor nodes start to collect and transmit data back to the
base station via multi-hop forwarding. Since each node does not know the exact position of
other nodes at the beginning, a self-organization protocol [3] is required to connect the sensor
nodes to form a communication network for data transmission.
The main objective of a routing algorithm is to find a better way for data transmission to save
electric power. Therefore, how to design an efficient routing algorithm to extend the lifetime of
WSN has become a very important research topic [4, 5]. LEACH is a classic routing protocol
[6] for cluster WSN because it can balance energy consumption within a cluster, and hence
extend the network lifetime. Its operation contains two stages: initial stage and stable stage. In
the initial stage, the base station selects a number of nodes as cluster heads based on random
thresholds. All other nodes join their nearby clusters by sending out signals and discover the
closest cluster heads. When clusters are formed, the wireless sensor network goes into a stable
stage. Each node starts to collect and transmit data to its cluster head, and the cluster head
forward these data to the base station together with their own collected data.
The reason for LEACH to randomly select a fixed number of cluster heads in each round is to
solve the problem of high energy consumption in direct-transmission approach and to balance
the power consumption among all sensor nodes. During the operation of each round, LEACH
generates a threshold value, which is sent to each sensor node to determine if it can become a
cluster head by a probability function. The new cluster head has to aggregate data from sensor
nodes in its cluster and transmit the data to the base station. In this way, all sensor nodes take
turns to serve as a cluster head to balance the power consumption. Many routing algorithms
were developed based on the same approach of LEACH in creating clusters, e.g., PAGASIS [7]
and TEEM [8].
PEGASIS works by connecting nodes in series, starting from the farthest node to the base
station, to form a linked structure with its neighbour nodes. When all nodes are connected, the
head of the linked structure will be selected. Then, every node combines both received data and
its own data and sends them towards the head. When all data arrive at the head, it will be
forwarded to the base station. Since the route is computed by the greedy algorithm, it is the
shortest and consumes less energy than LEACH. However, the greedy algorithm is complicated
and thus requires more energy in computation.
TEEN is also based on LEACH to transmit data to the base station periodically. It sets two
threshold values, i.e., hard threshold and soft threshold, to avoid the transmission of duplicated
data. This approach can save some energy by reducing the amount of data, but it is not suitable
for the applications requiring periodical data since the threshold values may not be met in
occasion. Therefore, a revised version of TEEN was proposed [9] to remedy the drawback of

52
 International Journal of Computer Science & Information Technology (IJCSIT) Vol 4, No 1, Feb 2012
reporting data periodically by reacting to sudden events in real time. Since both algorithms are
based on LEACH, the amount of energy saved is still limited.
The coverage ratio of a WSN represents the percentage of area covered by the working sensor
nodes among all application area, and it can also be used as an index for data integrity [10]. A
WSN can not achieve its function when the coverage ratio is too low. Besides, sensor nodes
have more feasible routes for selection to reduce energy consumption when the coverage ratio is
high. On the other hand, there often exist some holes [11] to block the shorter routes when the
coverage ratio is low, so the sensor nodes may have to use a longer route to bypass the holes for
data transmission. As the result, a lower coverage ratio may cause more energy consumption
and lower data integrity.
To extend the lifetime of WSN, a cluster allocation and routing algorithm based on node density
is proposed in this study. The objective is to solve the problem of LEACH and the related
routing algorithms for which the electric power of sensor nodes near the base station may
exhaust quickly and thus reduce the coverage ratio.
This study adopts a fixed cluster allocation mechanism and a deployment method based on the
node density for extending network lifetime. At the beginning, the algorithm allocates the
sensor nodes to a number of clusters, and the goal is to create clusters containing more sensor
nodes to share the energy consumption of their cluster head in forwarding data to the base
station.
For indirect transmission mode, the sensor nodes near the base station may exhaust their energy
quickly due to forwarding data for other sensor nodes, and it may also disable the data
transmission of the outer sensor nodes. Therefore, this study proposed to increase the node
density near the base station to compensate for the requirement of higher energy consumption in
this area. Without changing the total number of sensor nodes, the node density in the outer area
is reduced accordingly. The goal of the density deployment method is to remain a high coverage
ratio for extending the lifetime of WSN. Simulation results are provided to show the efficiency
of this approach.

2. RELATED METHODS
The cluster allocation and routing algorithm proposed in this study is based on LEACHs
clustered architecture. The difference is that the cluster allocation is done only once at the
beginning and remains fixed for the rest of time. Also, the node density near the base station is
increased during the deployment stage to achieve a higher coverage ratio and a longer lifetime
as well. The methods of high-energy-first mechanism for determining the cluster heads, finding
the forwarding node with the smallest angle [12], data compression [13] and sleep mode [14,
15] were also incorporated in the routing algorithm to further reduce the energy consumption
and extend the lifetime of WSN. These methods are described in the following.
A. High-Energy-First Method
In LEACH, a recently retired cluster head still has a chance to be selected again according to the
probability function, which may lead to fast exhaustion of its electric power. Therefore, this
study adopts the high-energy-first method to select cluster heads in each round to remedy this
drawback. After collecting data, each sensor node has to send out data together with the
information about its remaining electric power, and then the base station can decide which
nodes are to be selected as cluster heads in the next round using broadcast messages.
B. Determining Forwarding Node
In each round, a cluster head has to determine the forwarding node, which is also a cluster head
in its neighbourhood. The most direct forwarding route is determined based on the angle
between the directions of the base station and the forwarding node. The neighbour cluster head
53
 International Journal of Computer Science & Information Technology (IJCSIT) Vol 4, No 1, Feb 2012
with the smallest angle is chosen as the candidate first, and the purpose is to use the shortest
route to reduce energy consumption in forwarding data. However, if the cluster head with the
smallest angle refuses to forward data due to its low electricity, the one with the second smallest
angle is chosen next, and so on. If no cluster heads within 90 degree are available, the
requesting cluster head must transmit data directly to the base station when the distance is
reachable at the cost of consuming more energy; otherwise, it will seek for cluster heads outside
90-degree area as the forwarding node using a longer route.
C. Data Compression
In WSN, the amount of data transmitted can also affect the energy consumption of sensor nodes.
When a cluster head detects the same or similar data packets which are being transmitted, it can
use the data compression method to filter out similar data packets. This method can reduce the
amount of data and therefore save some energy.
D. Sleep Mode
When sensor nodes are scattered randomly during the deployment of a WSN, some nodes may
be too close to each other and collect the same data. Therefore, using sleep mode can reduce the
data amount and energy consumption. In this study, a grouping method is used to divide sensor
nodes into a number of groups within a cluster. In each group, the node with more electric
power is selected as the active node while the other nodes may enter sleep mode. The main
objective of the grouping method is to evenly distribute active nodes in each cluster to reduce
data similarity. Furthermore, the ratio of active nodes and the group size can be adjusted
according to the requirement of data precision.

3. CLUSTER ALLOCATION AND ROUTING

This study adopted a fixed cluster allocation algorithm similar to the initial stage of LEACH. At
first, the base station sends a threshold value to all sensor nodes for the selection of cluster
heads. Then, this study used the cluster allocation algorithm to create clusters with more sensor
nodes. After that, each sensor node sends the information back to the base station for later usage,
including node number, cluster number, location and remaining electricity. As soon as the
initialization stage is completed, the WSN begins its operation in each round to collect and
transmit data by the routing algorithm.

A. Cluster Allocation
The concept of set operation is used in dividing the sensor nodes in a WSN into a number of
clusters. At the beginning, the initial stage of LEACH is used for selecting cluster heads, and
then the cluster heads create their own clusters by communicating with the sensor nodes within
the sensing area. For simplicity, the created clusters are defined as the allocated sets, while the
sensor nodes not invited by any cluster heads are left in the unallocated set. Then, the allocated
sets are sorted by the number of their sensor nodes, and those in the intersected sets are re-
allocated to the set with more sensor nodes. The objective is to produce clusters with more
sensor nodes such that they are more powerful in forwarding data for other nodes. After that, the
sets with very few nodes will be deleted, and their nodes are put into the unallocated set.
When all sensor nodes in the intersected sets are re-allocated, each sensor node belongs to a
unique cluster. If there are still some sensor nodes in the unallocated set, the algorithm repeats
the same process as described above until the remaining sensor nodes have been allocated to a
certain cluster. The reason for re-allocating the sensor nodes in the intersected set is to produce
clusters with more sensor nodes to compensate for the high energy consumption by cluster
heads in forwarding data (Figure 1). The steps for cluster allocation are described in the
following:

54
 International Journal of Computer Science & Information Technology (IJCSIT) Vol 4, No 1, Feb 2012

Figure 1. Allocating the intersected set to the set with more nodes

Step1: The initial stage of LEACH is used for selecting cluster heads from all sensor nodes, and
then the cluster heads create a number of clusters by communicating with the sensor
nodes within the sensing area. The sensor nodes not invited by any cluster heads are
assigned in unallocated set C0.
Step2: Sort the allocated sets by their node numbers in descending order such that the number of
nodes in set Ci is greater than or equal to that of set Cj if i<j.

Step3: Starting from the allocated set with more nodes, assign the nodes in the intersection
Ci Cj to set Ci j i and i<j. If the assigned cluster head in set Cj Ci Cj , assign
the remaining nodes in Cj to C0 and remove Cj from the list of allocated sets.
Step4: If C0 , select some nodes in C0 as cluster heads to form more clusters. The nodes in
the newly created sets are removed from C0. Repeat step2 to step4 until C0 = or the
condition of the initial coverage ratio is satisfied.

B. Routing Algorithm
The routing algorithm operates by rounds. After sensor node deployment and cluster allocation,
the routing algorithm uses high-energy-first method to select sensor nodes with more electric
power as the cluster heads, and determines the forwarding nodes for data transmission. After
that, all sensor nodes start to collect data, which are aggregated by the cluster heads and
forwarded to the base station. Finally, the base station computes the coverage ratio to decide if
the operation will continue or not. If the coverage ratio is acceptable, the system goes to the next
round to repeat its operation. The cluster head decides its forwarding node by finding the
shortest route to reduce energy consumption. However, the neighbouring cluster heads may
refuse to forward data due to insufficient electric power. In that case, the cluster head initiates
the dynamic routing process to forward data by a feasible route when its forwarding request is
declined by other cluster heads.
As shown in Figure 2, cluster head A selects node 1 as its forwarding node initially, but is
declined due to low electric power; it then selects the node with the second smallest angle, i.e.,
node 4, but is declined again. Then, it selects the node with the third smallest angle, i.e., node 2,
but still in vain. The process continues until there are no forwarding nodes in the direction of
base station for selection. After that, the cluster head try to transmit data to the base station
directly if it is within the reachable range. Otherwise, it will turn to request nodes in the
direction greater than 90 degree as a last resort. When a cluster head has to request a
forwarding node deviating from the direction of the base station, it means most nodes in the
preferred direction are in low-power conditions. Instead of giving up at this moment, the
cluster head tries it best by using a detoured route to send out data since the goal is to maintain
a high receiving ratio by the base station.

55
 International Journal of Computer Science & Information Technology (IJCSIT) Vol 4, No 1, Feb 2012

Figure 2. Locating forwarding node in neighbouring clusters

4. SIMULATION RESULTS
This study conducted several simulation experiments to analyze if the proposed algorithm could
extend the lifetime of WSN. In addition, this study would like to investigate if raising the node
density around the base station without increasing the total number of sensor nodes can deal
with the high energy consumption problem in this area. This study also tried to find out a better
density proportion for usage in the deployment stage to extend the lifetime of WSN. The results
were compared with those of LEACH algorithm under the same conditions.
Before the simulation experiments, this study provides the following analysis about the power
consumption by sensor nodes. Basically, the energy required for transmitting a signal is highly
related to the distance [16]. The following equation shows the energy consumed when sending a
signal to a distance d by an amplifier.
2
Energy consumption = fs d , if d d 0
4
tr d , if d > d 0

Using d 0 as a threshold, if the transmission distance is shorter than d 0 , a free-space propagation

model is used to calculate the consumed energy, which is proportional to the square of distance.
If transmission distance is longer than d 0 , the two-ray ground propagation model is used for
calculation and the consumed energy is proportional to the fourth power of distance. In that
case, the consumed energy has a great influence on the wireless communication system. In the
above equation, fs and tr are the parameters for the free-space propagation model and two-
ray ground propagation model with their values equal to 10 pJ/bit/m 2 and 0.0013 pJ/bit/m 4 ,
respectively; Here, d 0 is defined as fs / tr , which is the threshold of transmission distance and
its value is about 87.7. To simplify the computation, it is assumed that d 0 =100m.

For most sensor nodes in WSN, the consumed energy is proportional to the square of distance
when collecting and sending data to their cluster head. The cluster allocation algorithm can
increase the node density in a cluster and thus reduce the distance and consumed energy in
transmitting data. As the operation continues, the sensor nodes near the base station may
exhaust their electricity and thus can not forward data for the outer sensor nodes. Consequently,
the outer sensor nodes may need to transmit data directly to the base station at a longer distance,
so the energy consumed is proportional to the fourth power of distance.
In this study, the simulation programs were written by Java. To obtain more accurate results,
all experiments were performed 100 times to compute the average results. The parameters for
the experiments are described in the following. The size of the application area is 700m700m

56
 International Journal of Computer Science & Information Technology (IJCSIT) Vol 4, No 1, Feb 2012
and there are 2000 sensor nodes deployed. The longest transmission distance for a senor node
is 200m. The initial energy in each sensor node equals 2J, and the energy for sending and
receiving data are the same, 50nJ/bit. Each data aggregation takes 5nJ/bit with the compression
rate = 70%. The amplifier parameters for the two models are fs =10pJ/bit/m2 and
4
tr =0.0013pJ/bit/m , respectively, and the data rate equals 160 bit/sec.

A. Computing Coverage Ratio

The coverage ratio of a WSN represents the percentage of area under monitoring (Figure 3),
which is computed as the area covered by the working sensor nodes divided by the total
application area (700m700m =490000m2). The sensor nodes can function correctly only when
they have enough electricity, so the electric power is the major factor affecting coverage ratio. In
this study, the WSN stops its operation if the coverage ratio goes below 80%, and the number of
rounds at this time is defined as its lifetime.

Figure 3. The area covered by working sensor nodes

B. Location of Base Station

The location of base station has a great influence on the energy consumption of sensor nodes
(Table 1). To find out a better location for the base station, an experiment was conducted to
simulate the operation of a WSN with the base station situated in three different locations, i.e.,
upper left (Figure 4), left (Figure 5), and center (Figure 6). In these figures, green spots indicate
the base station, and black lines represent the forwarding routes. The node colour is used to
represent the status of remaining electric power in a sensor node as follows: blue (sufficient),
red (low), and black (exhausted). The algorithm proposed in this study was used for collecting
and sending data, and the WSN stops its operation when the coverage ratio goes below 80%.
According to the lifetimes of the base station situated at three different locations, it can be seen
that the location of the base station plays an important role to affect the lifetime of WSN. The
lifetime is the longest when the base station is located at the center. The reason is that there
were more sensor nodes to share the high energy consumption around the base station.

57
International Journal of Computer Science & Information Technology (IJCSIT) Vol 4, No 1, Feb 2012
Table 1. Lifetime vs. the location of base station.

Location of base station Lifetime (rounds)

Upper left 1500
Left 2655
Center 8342

Figure 4. The base station is located at upper left

Figure 5. The base station is located at the left

58
 International Journal of Computer Science & Information Technology (IJCSIT) Vol 4, No 1, Feb 2012

Figure 6. The base station is located at the center

C. Comparison with LEACH

In this experiment, 2000 sensor nodes were deployed in a 700m700m application area. The
proposed algorithm and LEACH were used to collect and transmit data. The lifetimes (in
rounds) for both algorithms were recorded until the coverage ratio dropped below 80%. As we
can see in Figure 5, the initial coverage ratio is about 99% for both algorithms. For LEACH, the
coverage ratio goes below 80% at the time of 5002 rounds. For the proposed algorithm, the
coverage ratio goes below 80% at the time of 8342 rounds. Hence, the proposed algorithm can
efficiently increase the lifetime of WSN, and the percentage of increment is 66%.

Figure 7. Comparing the lifetimes of LEACH and proposed algorithm.

59
 International Journal of Computer Science & Information Technology (IJCSIT) Vol 4, No 1, Feb 2012
D. Adjusting Node Density
When the indirect transmission mode is used in WSN, the sensor nodes near the base station
often exhaust their energy first due to forwarding data for other nodes, which can also result in
the early death of WSN. As we can see from the above results, there are a lot of active sensor
nodes in outer areas not able to transmit data to the base station because most of the inner sensor
nodes have exhausted their electricity. Hence, this study proposed a density deployment method
to raise the density of sensor nodes near the base station without increasing the total number of
sensor nodes. The simulation results showed that this approach was very useful in extending the
lifetime of WSN.
To find out a better density proportion for deploying sensor nodes, an experiment was
conducted by dividing the application environment into 3 areas: (A) the area within 200m of the
base station, (B) the area of distance between 200m and 300m from the base station, and (C) the
area of distance longer than 300m from the base station. According to previous results, the
sensor nodes in area A will consume more energy, and those in area B will consume less
energy, and those in area C will consume the least energy.
By adjusting the node density in these 3 areas, this study tried to find out a better density
proportion to increase the coverage ratio of WSN and its lifetime as well. There are 2000 sensor
nodes deployed according to the designated density in each area. Because the proportion of
these 3 areas is 2002 : (300 2 200 2 ) : (700 2 300 2 ) =4 : 5 : 49-9, if their node densities are
the same, the numbers of sensor nodes in these three areas are computed as NA=2000 4 =513,
49
NB=2000 5 =641, and NC=2000 (49 9 ) =846.
49 49
The experiment compared the lifetimes (in rounds) for five different density proportions when
the coverage ratio dropped below 80% by the proposed algorithm and LEACH. It is noted that
the lifetime for each case was computed by the average of 100 simulation results. The results in
Table 2 shows that raising the node density in the area near the base station can increase the
lifetime of WSN. When the density proportion is 4:2.5:1, the lifetime for the cluster allocation
and routing algorithm is the longest (16789 rounds). The same situation can also be found in
LEACH algorithm, and the longest lifetime equals 11379 rounds when the density proportion is
4:2.5:1.

Table 2. Lifetime for different density proportions (the proposed algorithm and LEACH)

Node number Rounds when coverage ratio dropped below 80%

Density proportion
A B C Proposed Algorithm LEACH
111 513 641 846 8342 5002
21.51 724 678 598 13569 9932
321 839 699 462 15073 10622
42.51 912 712 376 16789 11379
531 961 721 318 15598 10676
63.51 998 727 275 16288 11019

As we can see from the above results, raising the node density near the base station can increase
the lifetime of a WSN. This study further compared the results of LEACH and the proposed
algorithm under the density proportion of 4:2.5:1. As shown in Figure 8, the initial coverage
ratios for both algorithms are about 95%. The coverage ratio of LEACH drops to 80% at the
time of 11379 rounds, which is increased by 127% compared to the case without adjusting node
density (5002 rounds). The coverage ratio of the proposed algorithm drops to 80% at the time of
60
 International Journal of Computer Science & Information Technology (IJCSIT) Vol 4, No 1, Feb 2012
16789 rounds, increased by 101% compared to the case without adjusting node density (8342
rounds). For both algorithms, their lifetimes have been increased by more than 100% (as shown
in Table 3) if the proposed density deployment method was applied.

Figure 8. Comparing the results of both algorithms after density adjustment

Table 3. Initial coverage ratio and lifetime by adjusting node density in both algorithms

Rounds when coverage ratio

Initial coverage ratio
dropped below 80%
Algorithm
Density adjusted Density adjusted
No Yes No Yes
LEECH 99% 95% 5002 11379
proposed algorithm 99% 95% 8342 16789

5. CONCLUSIONS
How to deploy sensor nodes to cooperate with efficient routing algorithms for extending
lifetime is an important topic in the research of WSN. For large-area applications, most routing
algorithms adopt indirect transmission mode to solve the problem of high energy consumption
due to long-distance transmission, but it may also cause the sensor nodes near the base station to
exhaust electricity quickly. This study proposed a cluster allocation and routing algorithm based
on node density to remain a high coverage ratio and thus extend the lifetime of WSN. The
simulation results showed that the lifetime (in rounds) for the proposed algorithm was 66%
higher than that of LEACH.
In addition, this study proposed a deployment method to adjust node density without increasing
the total number of sensor nodes. The objective was to increase the sensor nodes near the base
station to compensate for the high energy consumption due to forwarding data for other nodes.
The simulation results showed that it was very efficient in increasing the lifetime of WSN. For
the proposed algorithm and LEACH, their lifetimes were increased by more than 100% if the
density proportion of 4:2.5:1 was used for node deployment. Therefore, the cluster allocation

61
 International Journal of Computer Science & Information Technology (IJCSIT) Vol 4, No 1, Feb 2012
and routing algorithm as well as the density deployment method proposed in this study are very
efficient in extending the lifetime of WSN.

ACKNOWLEDGEMENTS
The authors would like to thank for the financial support of the National Science Council (NSC)
in Taiwan under the contract numbers NSC 100-2511-S-134-003.

REFERENCES
[1] I. F. Akyildiz, W. Su, Y. Sankarasubramaniam, and E. Cayirci, Wireless Sensor Networks: a
Survey, Computer Networks, vol. 38, no. 4, pp. 393-422, 2002.
[2] W. R. Heinzelman, A. Chandrakasan, and H. Balakrishnan, Energy-efficient communication
protocol for wireless microsensor networks, Proceedings of the 33rd Hawaii International
Conference on System Sciences, vol. 8, pp. 8020, Jan. 4-7, 2000.
[3] K. Sohrabi, J. Gao, V. Ailawadhi, and G. J. Pottie, Protocols for self-organization of a wireless
sensor network, IEEE Personal Communications, vol. 7, no. 5, pp. 16-27, 2000.
[4] Harish. H. Kenchannavar, Sushma. S. Kudtarkar, U. P. Kulkarni, Energy Efficient Data Processing
in Visual Sensor Network, International Journal of Computer Science and Information
Technology, vol.2, no.5, pp. 151-160, 2010.
[5] Laiali Almazaydeh, Eman Abdelfattah, Manal Al-Bzoor, and Amer Al-Rahayfeh, Performance
Evaluation of Routing Protocols in Wireless Sensor Networks, International Journal of Computer
Science & Information Technology, vol.2, no.2, pp. 64-73, 2010.
[6]
M. J. Handy, M. Haase and D. Timmermann, LEACH-C Low energy adaptive clustering
hierarchy with deterministic cluster-head selection, 4th IEEE International Conference on Mobile
and Wireless Communication Network, pp. 368372, Sep. 2002.
[7] S. Lindsey and C. S. Raghavendra, PEGASIS: Power efficient gathering in sensor information
systems, Proceedings of the IEEE Aerospace Conference, vol. 3, pp. 1125-1130, 2002.
[8] A. Manjeshwar and D. P. Agrawal, TEEN: a routing protocol for enhanced efficiency in wireless
sensor networks, Parallel and Distributed Processing Symposium, pp. 2009-2015, Apr. 2001.
[9] A. Manjeshwar and D. P. Agrawal, APTEEN: a hybrid protocol for efficient routing and
comprehensive information retrieval in wireless sensor networks, Proceedings of the International
Parallel and Distributed Processing Symposium, pp. 195-202, 2002.
[10] S. Meguerdichian, F. Koushanfar, M. Potkonjak and M. Srivastava, Coverage problems in wireless
ad-hoc sensor networks, Proceedings of IEEE INFOCOM vol. 3, pp. 1380-1387, Apr. 2001.
[11] R. Ghrist and A. Muhammad, Coverage and hole-detection in sensor networks via homology,
Proceedings of the 4th International Symposium on Information Processing in Sensor Networks,
pp. 254-260, Apr. 24-27, 2005.
[12] Wernhuar Tarng, Kuo-Liang Ou, Kun-Jie Huang, Li-Zhong Deng, Hao-Wei Lin, Chang Wu Yu,
Kuen-Rong Hsieh and Mingteh Chen, Applying Cluster Merging and Dynamic Routing
Mechanisms to Extend the Lifetime of Wireless Sensor Networks, International Journal of
Communication Networks and Information Security, vol. 3, no. 1, pp. 8-16, 2011.
[13] X. Tang and J. Xu, Extending network lifetime for precision-constrained data aggregation in
wireless sensor networks, Proceedings of IEEE INFOCOM, Apr. 2006.
[14] J. M. S. N., L. B. Ruiz and A. F. Loureiro, Manna: A management architecture for wireless sensor
networks, IEEE Communication Magazine, vol. 41, pp. 116125, 2003.
[15] V. Raghunathan, C. Schurgers, S. Park, and M. B. Srivastava, Energy-aware wireless micro sensor
networks, IEEE Signal Processing Magazine, vol. 19, no. 2, pp. 40-50, 2002.
[16] W. B. Heinzelman, P. Chandrakasan, and H. Balakrishnan, An application-specific protocol
architecture for wireless micro sensor networks, IEEE Transactions on Wireless Communications,
vol. 1, no. 4, pp. 660-670, 2002.

62