Energy Balanced Clustering Algorithm in Large Scope WSN ...
Transcript of Energy Balanced Clustering Algorithm in Large Scope WSN ...
Energy Balanced Clustering Algorithm in Large Scope
WSN Based on Cooperative Transmission Technology
Abstract—In this paper, a clustering algorithm based on the
cooperative transmission technology is proposed with respect to
uneven energy consumption in large scope wireless sensor
network. In this algorithm, to ensure the even distribution of the
cluster head nodes in large scope wireless sensor network which
has even clusters, the cooperative transmission technology is
applied in data transmission among the clusters. System
performance is improved by virtual multi-antenna system which
is constituted through cooperative transmission between the
sensor nodes. The experimental results show that the energy
consumption is balanced and the network lifetime is extended. Index Terms—Wireless sensor network, energy consumption
unbalanced, clustering algorithm, cooperative transmission,
cluster head node
I.
INTRODUCTION
Wireless sensor network (WSN) is a self-monitoring
system through MANET way, which is composed of a
large number of sensor nodes deployed in the monitoring
area. It has been gradually used in military, civilian and
industrial areas. However, due to the huge number of
sensor nodes and the general use of battery as the power
source, the energy issue has become a bottleneck for
further development.
As multi-hop manner is generally used in data
transmission of the wide range of WSNs, the loads of the
nodes near the SINK will be inevitably increased and
more energy is consumed due to transponding the data of
the peripheral nodes. It results in the uneven energy
consumption of the entire network. In papers [1]-[4] the
relay nodes, which can transpond data but do not perceive
information, are added to the network, thus avoiding the
problem of uneven energy consumption caused by
excessive load. However, a huge number of relay nodes
are added in [4], which is not applicable in practice. Then
an energy consumption balancing strategy based on
adjustable transmit power is proposed in [5]. When the
residual energy of the node is high, the transmitting
power is increased and the data is transmitted to longer
distance; otherwise, the transmitting power will be
reduced [6]. A co-operative transmission strategy based
on residual energy stimulation is proposed in literature
This work is supported by the Major National Scientific and
Technological Project of China (No. 1010122720110046). Corresponding author email: [email protected]
[7]. When the co-operation conditions are met, data will
be sent to the SINK through the cooperative transmission
node. To some extent, it solves the problem of uneven
energy consumption in the network. However, the
method is not applicable to large scope WSNs. The
energy consumption cannot reduce linearly as the number
of cooperative nodes increase. When over a certain range,
the increase in the number of cooperative nodes cannot
improve the performance of the network significantly [8],
[9]. George N. Bravos did a deep research for
collaborative MIMO and further demonstrated the
cooperative MIMO communication has better energy
performance than the multi-hop SISO algorithm [10].
Based on the current status of research it can be seen
that clustering can solve the problem of local uneven
energy, and that the cooperative transmission can oppose
the multipath fading effects [11]. The combination of
them can effectively solve the problem of the uneven
energy consumption. For example, the LEACH routing
protocol requires that the cluster head nodes can
communicate directly with the SINK node. However,
according to the energy consumption model of the
wireless transmission, the energy consumption of the
wireless communication is proportional to the n-th power
of the transmission distance (n = 2, 3, 4). Thus, the local
energy balance can be ensured by the LEACH routing
protocol cluster, but the uneven energy problem brought
by the distance cannot be resolved [12]. If the
transmission technology is collaborated, such uneven
energy problem of distances can be solved. This paper is
studied based on this combination.
In this paper, a clustering algorithm based on the
cooperative transmission technology is proposed. With
such a method, the energy consumption in large scope
wireless sensor network is balanced and the network
lifetime is extended.
The rest of this paper is organized as follows: Section
II presents some foundations and the model for the
wireless sensor network system including the energy
consumption is discussed. In Section III, the proposed
methods for constitution of clusters, the determination of
the optimal cluster head nodes and the inter-cluster
communication strategy based on cooperative
transmission are described. Section IV presents the
simulation parameters and results. Finally, Section V
concludes the paper.
627
Journal of Communications Vol. 9, No. 8, August 2014
©2014 Engineering and Technology Publishing
doi:10.12720/jcm.9.8.627-633
Manuscript received March 12, 2014; revised August 27, 2014.
Yang Jiang, Yiqian Yan, and Yang Yu
College of Communication Engineering, Chongqing University, Chongqing 400044, China
Email: {cqjiangsun, yanyiqiancq}@126.com; [email protected]
II. SYSTEM MODEL
Since the wireless sensor networks use the clustering
algorithm, the network model is a hierarchical structure,
as shown in Fig. 1.
SINK
Head node
Member node
Fig. 1. The model of the network
For this model, the following assumptions are made: (a)
Once deployed, the nodes in the network are stationary;
(b) Except the SINK nodes have no energy limitation, the
other nodes in the network have the same initial energy;
(c) The nodes in the network can automatically adjust the
transmitting power.
The common energy model of wireless sensor net-
works is as follows.
_ _
2
0
4
0
( , ) ( ) ( , )
,
,
Tx Tx elec Tx amp
elec fs
elec amp
E k d E k E k d
kE k d d d
kE k d d d
(1)
_( ) ( )
Rx Rx elec elecE k E k kE (2)
where ( , )Tx
E k d is the energy consumed when the sensor
node transmits k bits of data, _ ( )Tx elecE k is the energy
consumed by the transmitting circuit when the sensor
node transmits k bits of data, _ ( , )Tx ampE k d is the energy
consumed by the amplifier when the sensor node
transmits k bits of data, ( )RxE k is the energy consumption
when the node receives k bits of data, and only the
energy consumption of the circuit is considered and elecE
indicates the loss energy of the transmitting circuit. If the
transmission distance is less than the threshold value 0d ,
the loss of the power amplifier uses the free space loss
model. Otherwise, the multi-path fading model is used.
fs and amp are the energies that the power amplifier
required for the free space model and the multi-path
fading model, respectively.
In the literature [13], for vMISO system, it is assumed
that the synergistic encode is STBC and the modulation
method is BPSK. When the bit error rate is the 10-3
, the
multi-path fading parameter =3 , the diversity gain,
G(Nc), under different cooperative node numbers, Nc,
and range extension, βext, when the nodes use the same
transmitting power are shown in Table I[10]
.
TABLE I: DIVERSITY GAIN AND RANGE EXTENSION
Nc 2 3 4 5 10
G(Nc)(dB) 10 13.5 14 14.5 15.9
ext ( =3) 2.71 4.07 4.65 5.2 7.3
The number of cooperative nodes can be determined
according to the parameters in Table I. When
2.71link linkd d d , two cooperative nodes can transmit
data to the SINK node. And if 2.71 4.07link linkd d d ,
three nodes are needed to accomplish the transmission.
The impact of the cooperative transmission on the
transmission range is described in Table I. Assume that
the transmission distance is constant, with the cooperative
transmission added the transmitting power of the nodes
will inevitably decline. The transmitting power of the
node after collaboration is shown in
( )10
10 cG N
c
t tP P (3)
where c
tP represents the transmitting power of a single
node with cooperative transmission, and Pt represents
that without cooperative transmission. Thus, the equation
(3) describes the relationship of the transmitting power
with and without cooperative transmission.
If Rs
is the energy consumed when the sensor node
transmits s bits of data and the transmitting time is 1s,
the energy model is as follows.
2
0
4
0
* * ,
* * ,
s elec fs
s elec amp
E Rs E Rs d d d
E Rs E Rs d d d
(4)
According to (3) and (4), when the number of
cooperative nodes is tM , the consumption per second
can be present as
2 10
0
4 10
0
* * * *10 ,
* * * *10 ,
c
c
G N
t elec fs
G N
t elec amp
E M Rs E Rs d d d
E M Rs E Rs d d d
(5)
Thus, we get the mathematical model of the energy
consumption with the cooperative transmission.
0 20 40 60 80 100 120 140 160 180 2000
0.002
0.004
0.006
0.008
0.01
0.012
0.014
0.016
0.018
the distance from SINK /m
the e
nerg
y c
onsum
ption o
f th
e n
odes/J
no cooperative nodes
two cooperative nodes
three cooperative nodes
four cooperative nodes
Fig. 2. relationship between energy consumption and the number of
cooperative nodes
628
Journal of Communications Vol. 9, No. 8, August 2014
©2014 Engineering and Technology Publishing
Fig. 2 shows the relationship between energy
consumption and the number of cooperative nodes. In the
case of the same transmitting distance, the cooperative
nodes can relatively reduce the energy consumption of
the nodes. However, the energy consumption is not
always less as the number of cooperative nodes is more.
From Fig. 2 it can be seen that, the energy consumption
of the four cooperative nodes is greater than that of three
cooperative nodes.
III. EBBMCC_LS CLUSTERING ALGORITHM
A. Constitution of Clusters
1) Constitution of the initial uniform clusters
When all the nodes are deployed to the monitoring area,
the nodes obtain their location by initialization. It is
assumed that the location of SINK is (x0,y0) and the
number of sensor nodes is n. When the nodes are
deployed, their locations are fixed and expressed as
[(x1,y1),(x2,y2)….(xi,yi)….(xn,yn)]. The nodes calculate
their distances from the SINK d as
2 2
0 0( ) , 1,2,..... .i id x x y y i n (6)
If 1 2d r A n , select the node as the candidate
cluster head node, where r is the radius of the cluster
domain and A is the area of the network. Assume there
are m candidate cluster head nodes. The candidates
broadcast their identities. Then the nodes received the
broadcasted information report their locations to the
candidates. The candidates calculate their distances from
these nodes as
2
2( ) , =1,2,...... k k j k jd x x y y k m (7)
where j is the number of the nodes which have reported
their locations. If 02 kr d d , node j will be selected as
a candidate. And the marked candidates which do not
satisfy the conditions will be canceled. Thus, each
candidate performs the operation to find the nearby
candidates, until the whole network is covered. Then the
candidates distribute uniformly in the network.
When the candidates are confirmed, the candidates
broadcast their identities with the level of the transmitting
power corresponding to the radius of the cluster domain.
The nodes received this message join the cluster, which
forms an initial uniformly distributed cluster structure.
The flow chart of building initial cluster is shown in Fig.
3.
2) Determine the optimal cluster head nodes
Compared to other nodes, the cluster head nodes
consume more energy, so the nodes with more residual
energy should be selected. While in order to minimize
energy consumption, the nodes should have minimum
communication cost. The initial cluster structure is just
uniformly distributed without considering the problem of
energy consumption. Therefore, the cluster head must be
optimized.
Start
Deploy n sensor nodes
The Nodes
initialize and get
their locations
The nodes calculate their distances from the SINK
1
2An
d r
2 2
0 0( )i id x x y y
be the candidate cluster
head node and join in
The candidates calculate their distances from node j
2
2( )k k j k jd x x y y
kd
12
2A
k nd r
Add node j to
Delete node j from
headS
headS
headS
is traversed headSk=k+1
RNHm is traversed
( )NHmR
j=j+1
Yes
Yes
i>n?i=i+1No
Yes
No
No
Broadcast their identities as the
candidates
Joining the
corresponding clusters
End
No
Yes
No
Yes
Fig. 3. The flow chart of building initial cluster
The minimum communication cost of the cluster head
node j is calculated as follows.
min
1
1cos
m
j k
k
t Pm
(8)
where m is the number of cluster members and Pk is the
629
Journal of Communications Vol. 9, No. 8, August 2014
©2014 Engineering and Technology Publishing
power level that the cluster member k uses to send data.
The above equation represents the average trans-
mitting power when all the cluster members send data to
the cluster head node. With the consideration of the
residual energy level, the minimum communication cost
is
_ min
1cos _ cos
( )j i
req
t adjusted tE j
(9)
where ( )reqE j is the residual energy of the j-th sensor
node.
According to the amended smallest cluster
communication cost, the optimal cluster nodes are
determined as the cluster head nodes. Thus, a single
round cluster is built. At this time, it starts the data
transfer phase for data sensing and transmission. After a
round ends, the optimization process of the cluster heads
repeats until the network dies.
After the optimal cluster head nodes are confirmed,
they broadcast their identities to the cluster members.
Thus, uniformly clustering phase in the network is
completed. The network enters the data transfer phase.
Data transmission in WSNs includes transmission in a
cluster and between clusters. Considering the impact of
energy consumption, the clustering range can meet the
free space loss model. Thus, each member node sends
data directly to the cluster head node in its own slot. The
algorithm aims at large-scale wireless sensor networks, so
data transmission between clusters is much more complex.
The flow chart of confirming best cluster head is shown
in Fig. 4.
j=1
i=1
_ min
( )
1cos _ cos
i i
req j
t adjusted tE
j=m?(m is the number of cluster
members)
Select the node with
smallest
as the head nodecos _
it adjusted
i=n?
(n clusters)
j=j+1
i=i+1
Yes
End
No
No
Yes
Fig. 4. The flow chart of confirming best cluster head
B. Inter-Cluster Communication Strategy Based on
Cooperative Transmission
In general, the cluster head can send integrated
information to the base station with the following two
means. One is the single hop communication method, that
is, the fusion of data is sent to the remote base station
directly. Another is obtaining the path of the minimum
energy cost from the cluster head to the base station and
realizing multi-hop data transmission by relaying and
forwarding of the other cluster head nodes. The first
method cannot meet the needs of energy conservation.
The other one with a multi-hop manner can effectively
reduce the number of cluster head nodes which directly
communicate with the base station. The data can be
transmitted by the cluster head close to the base station.
Although the algorithm is more complex, the study
focuses on the second communication method because of
the energy efficiency of WSN. Therefore, the proposed
EBBMCC_LS (Energy Balanced Based Multi-hop and
Cooperative transmission technology in Clustered Large
Scope WSN) clustering algorithm combines the initial
multi-hop with cooperative transmission.
The key issue of inter-cluster communication is the
problem of cooperative transmission.
1) Conditions of cooperative transmission
Assume CHnT and CHmR are the two cluster head nodes
of multi-hop communication, in which CHnT is the
sending node, where the average residual energy of the
cluster is re nTE CH and CHmR is the receiving node,
where the average residual energy of the cluster is
re mRE CH .
If re nT re mRE CH E CH , the cooperative trans-
mission is used for the expansion of the transmission
distance; if re nT re mRE CH E CH , the initial multi-hop
transmission is used.
2) Selection of cooperative nodes
If re nT re mRE CH E CH , the sending node uses
cooperative transmission to extend the transmission
distance, thus avoiding the traffic load of the next hop.
The cooperative node is selected from the collection RNHm.
According to Table I, when the number of cooperative
nodes is 2, 3 or 4, and single transmission power is Pt, the
extension is 2.71, 4.07 or 4.65, which means that the
expansion is good. If the cooperative nodes are more than
4, the effect of the expansion is significantly reduced. In
order to better reflect the expansion characteristics of the
cooperative transmission, only the cases of 2, 3 and 4
cooperative nodes are considered. When condition of
cooperative transmission is met and there is data to be
sent, according to the principle of a minimum number of
hops, the number of cooperative nodes can be determined
by the following equation
min , 2,3,4. s
ci
di
d
(10)
where ds represents the distance of the sending node from
the SINK node and dci represents the transmitting range
when the number of nodes is i.
To maximizing the energy in the cluster after the data
is sent, the cooperative nodes have to satisfy the two
conditions: a) the residual energy of the nodes is as much
630
Journal of Communications Vol. 9, No. 8, August 2014
©2014 Engineering and Technology Publishing
as possible. Because more energy will be consumed in
the collaborative process; b) the distances of the nodes
from the cluster head node need to be as small as possible
to avoid the extra energy consumption.
Thus, select the sensor nodes meeting the following
condition as the cooperative nodes:
max( ), 1,, , ,
re i
ci
E CHi n
d (11)
where re iE CH is the residual energy of the i-th node,
cid is the distance from the i-th node to the cluster head
node and n is the number of cluster members. The flow
chart of communication among clusters is shown in Fig. 5.
Determine the next hop of
the cluster head by The
way of simple multi-hop
>req tE req rE
Send data to the
next hop of the
original multi-hop
Choose as the
cooperative node
Cooperative transmission
max( )
re i
ci
E CH
d
Yes
No
Fig. 5. The flow chart of communication among clusters
IV. SIMULATION
A. Simulation Settings
MATLAB is used as the simulation platform. The
energy loss of the network is calculated with the model of
radio communication system. The parameters in the
simulation experiment are shown in Table II. In the
simulation, when the node energy is less than Emin
(0.002J), the node is considered to be death.
TABLE II: SIMULATION PARAMETERS
parameters value
Network Range (300×300) , (400×400) , (500×500)
Number of nodes 300, 500, 700 Location of base station (150,150), (200,200), (250,250)
Initial energy of node 0.5J
Radius of cluster 30m Eelec 50nJ/bit
Efs 10pJ/bit/m2 Emp 0.0013pJ/bit/m4
Threshold(d0) 75m
B. Distribution of the Cluster
Clustering constitutes the hierarchical structure of
WSN. The energy consumption of the network can be
better balanced with the cyclical cluster head rotation
strategy. Meanwhile, the clustering algorithm also has its
own shortcomings, such as the energy consumption is
only balanced at a local scale for uneven clustering, etc.
Especially in large-scale wireless sensor networks, due to
inter-cluster data forwarding, uneven energy consumption
is caused inevitably by the distance. Simple clustering
algorithm is not an effective solution. EBBMCC_LS
algorithm is a clustering routing algorithm to balance the
energy consumption of a wide range of wireless sensor
network. The distribution of the clusters is verified first.
0 50 100 150 200 250 3000
50
100
150
200
250
300
Fig. 6. Distribution of nodes
Fig. 6 is the distribution of 300 sensor nodes randomly
distributed in a square area of 300 * 300 m2. The nodes
are deployed randomly, so the distribution of them is
uneven as in the figure.
After running the clustering algorithm to build clusters
based on the node deployment in Fig. 6, the resulted
network is as in Fig. 7. The dark nodes represent the
cluster head nodes. From the figure, the cluster head
nodes are relatively evenly distributed in the network.
Therefore, EBBMCC_LS algorithm can achieve a
uniform distribution of nodes in the network clustering.
0 50 100 150 200 250 3000
50
100
150
200
250
300
Fig. 7. Distribution of initial cluster heads in WSN
C. Analysis of Energy Consumption
This study focuses on the problem of energy
consumption of wireless sensor networks. Now the
energy consumption of EBBMCC_LS algorithm is
evaluated by deaths of nodes. If all nodes die in a short
631
Journal of Communications Vol. 9, No. 8, August 2014
©2014 Engineering and Technology Publishing
period, it indicates that the energy consumption of all
nodes in the network is relatively even; otherwise, the
energy consumption of the network is seriously uneven.
0 200 400 600 800 1000 1200 1400 1600 1800 20000
50
100
150
200
250
300
回合数/轮
死亡节点数
/个
LEACH
LEACH(mhop)
LEACH-C(mhop)
EBBMCC-LS
Th
e n
um
be
r o
f th
e d
ea
d n
od
es
The round of the network Fig. 8. Dead nodes in 300*300 m2 network areas
0 200 400 600 800 1000 1200 1400 1600 1800 20000
50
100
150
200
250
300
350
400
450
500
回合数/轮
死亡节点数
/个
LEACH
LEACH(mhop)
LEACH-C(mhop)
EBBMCC-LS
The
nu
mb
er o
f th
e d
ead
no
des
The round of the network
Fig. 9. Dead nodes in 400*400 m2 areas
0 200 400 600 800 1000 1200 1400 1600 1800 20000
100
200
300
400
500
600
700
回合数/轮
死亡节点数
/个
LEACH
LEACH(mhop)
LEACH-C(mhop)
EBBMCC-LS
The
nu
mb
er
of
the
de
ad n
od
es
The round of the network
Fig. 10. Dead nodes in 500*500 m2 network areas
Fig. 8, Fig. 9, Fig. 10 show the number of dead nodes
with the change of the number of rounds in LEACH
clustering single-hop transmission, LEACH clustering
multi-hop transmission, LEACH-C clustering multi-hop
transmission and EBBMCC_LS algorithm. The network
ranges are 300300 m2, 400400 m
2, 500500 m
2,
respectively. From the three figures, it can be seen that
death nodes appears firstly under LEACH clustering
single-hop transmission, and death of all nodes requires
the longest time. Then the LEACH clustering multi-hop
transmission follows. The next is LEACH-C clustering
multi-hop transmission. For EBBMCC_LS algorithm
death nodes appear the latest, and all the nodes die
quickly after the first death node. By comparing the
LEACH clustering single-hop transmission with the
LEACH clustering multi-hop transmission, it can be seen
that the single-hop approach is no longer applicable in the
wide range of wireless sensor networks; from the
comparison of LEACH clustering multi-hop transmission
and the LEACH-C clustering multi-hop transmission, a
uniform distribution of nodes can better balance the
energy consumption. Therefore, it is obvious that
EBBMCC_LS algorithm has better energy consumption
equalization.
D. Analysis of Network Lifetime
The lifetime of wireless sensor network is generally
defined as the moment when the first node dies, called
FND (First Node Died) Lifetime. Network lifetime is one
of the key measures of wireless sensor networks. In three
different network ranges of 300300 m2, 400400 m
2,
500500 m2, the improvements of the network lifetime
with EBBMCC_LS algorithm are verified by comparing
the time of the first dead node with the LEACH
clustering single-hop transmission, LEACH clustering
multi-hop transmission, LEACH-C clustering multi-hop
transmission.
300*300 400*400 500*5000
500
1000
1500
2000
网络范围/m*m
第一个死亡节点出现轮数
/轮
LEACH
LEACH(mhop)
LEACH-C(mhop)
EBBMCC-LS
The
mo
men
t o
f th
e fi
rst
no
de
die
s
The range of the network/m*m
Fig. 11. Comparison of network lifetime
Fig. 11 represents the comparison of network lifetime
under the above four algorithms in different ranges. In the
range of 300 * 300 m2, the lifetime of the EBBMCC_LS
algorithm is 1.4 times larger than the LEACH-C multi-
hop transmission, 1.8 times larger than the LEACH
multi-hop the transmission and 4.2 times larger than the
LEACH single-hop transmission. In the range of 400 *
400 m2, the lifetime of the EBBMCC_LS algorithm is
respectively 1.7 times, 2.2 times and 9 times than the
other methods. And in the range of 500 * 500 m2, the
lifetime of the EBBMCC_LS algorithm is respectively
1.9 times, 2.7 times and 23 times than the other methods.
It can be seen that, with the increase of network range,
improvements in the network lifetime of EBBMCC_LS
algorithm are increasingly apparent.
632
Journal of Communications Vol. 9, No. 8, August 2014
©2014 Engineering and Technology Publishing
According to the above simulation and analysis about
distribution of nodes, death of nodes and network lifetime
of the four algorithms, EBBMCC_LS algorithm has a
good capability of energy consumption equalization in a
wide range of wireless sensor networks, and is able to
postpone the first death of the nodes to extend the FND of
the network.
V. CONCLUSIONS
A clustering algorithm based on the cooperative
transmission technology with respect to uneven energy
consumption in large scope wireless sensor network has
been proposed. In this algorithm, the cooperative
transmission technology has been applied in data
transmission among the clusters, which has ensured the
even distribution of the cluster head nodes in large scope
wireless sensor network. It has been shown that the
energy consumption is balanced and the network lifetime
is extended. The performance of the network is improved
significantly.
REFERENCES
[1] S.Olariu and I. Stojmenovic, “Data-centric protocols for wireless
sensor networks,” Handbook of Sensor Networks: Algorithms and
Architectures, I. Stojmenovic, Ed., Wiley, 2005, pp. 417-456.
[2] G. Wang, L. S. Huang, H. L. Xu and Y. Wang, “Reliable relay
node placement in wireless sensor network,” in Communication
and Networking, China, 2008, pp. 371-375.
[3] G. Wang, L. S. Huang, H. L. Xu and J. B. Li. “Relay node
placement for maximizing network lifetime in wireless sensor
networks,” in Wireless Communications, Networking and Mobile
Computing, China, 2008, pp. 1-5.
[4] Y. F. Xin, T. Guven, and M. Shayman, “Relay deployment and
power control for lifetime elongation in sensor networks,” in Proc.
IEEE ICC, 2006.
[5] D. Li, X. L. Wu, and L. Yang, “Energy-efficient dynamic
cooperative virtual MIMO based routing protocol in wireless
sensor networks,” in Proc. 8th International Conference on
Communications and Networking in China (CHINACOM), 2013.
[6] T. Q. S. Quek, M. Z. Win, and D. Dardari. “Energy efficiency of
dense wireless sensor networks: To cooperate or not to cooperate,”
IEEE Journal on Selected Areas in Communications, vol. 25, no. 2,
pp. 459-470, 2007
[7] J. W. Jung and M. A. Ingram, “Residual-energy-activated
cooperative transmission (REACT) to avoid the energy hole,” in
Proc. IEEE International Conference on Communications
Workshop on Cooperative and Cognitive Mobile Networks, May
2010.
[8] V. K. Sachan, S. A. Imam, and M. T. Beg, “Energy-efficiency of
virtual cooperative MIMO techniques in wireless sensor
networks,” in Proc. International Conference on Computer
Communication and Informatics, Jan., 2012.
[9] R. M. Bani Hani and A. A. Ijjeh, "A survey on LEACH-based
energy aware protocols for wireless sensor networks," Journal of
Communications, vol. 8, no. 3, pp. 192-206, 2013.
[10] G. N. Bravos and A. G. Kanatas, “MIMO-based and SISO
multihop sensor networks: Energy efficiency evaluation,” Wireless
and Mobile Computing, Networking and Communication, pp. 13,
Oct. 2007.
[11] G. H. Chen, C. F. Li, M. Ye, and J. Wu, “An unequal cluster-
based routing strategy in wireless sensor networks,” Wireless
Networks (JS), vol. 15, no. 2, pp. 193-207, February 2009.
[12] W. R. Heinzelman, A. Chandrakasan, and H. Balakrishnan,
“Energy-efficient communication protocol for wireless
microsensor networks” in Proc. 33rd Hawaii Int’1.Conf. Sys. Sci,
Jan. 2000.
[13] T. D. Nguyen, O. Berder, and O. Sentieys, “Cooperative MIMO
schemes optimal selection for wireless sensor networks,” in Proc.
Spring Vehicular Technology Conference, Dublin, Ireland, 2007,
pp. 85-89
633
Journal of Communications Vol. 9, No. 8, August 2014
©2014 Engineering and Technology Publishing
Yang Jiang was born in Sichuan Province,
China, in 1963. He received his first class
honors B.S. degree in electronics from
Chongqing University in 1987, and his M.S.
and Ph.D. degrees both from Chongqing
University in 2000 and 2011, respectively.
Since 1987, he has been with the Department
of Electronic Engineering in Chongqing
University, where he is currently a professor.
His current research interest is in wireless communications,
communication systems and Internet of Things technology.
Yiqian Yan was born in Hebei Province,
China, in 1991. She received the B.S. degree
from Chongqing University, in 2012 in
Communication Engineering. She is currently
pursuing the M.S. degree with the College of
Communication Engineering, Chongqing
University. Her research interest is in wireless
communications, especially in wireless sensor
networks.
Yang Yu was born in Zhejiang Province,
China, in 1989. He received the B.S. degree
from Chongqing University, in 2012 in
Communication Engineering. He is currently
pursuing the M.S. degree with the College of
Communication Engineering, Chongqing
University. His research interest is in
Electronics and Communications.