Wireless Sensor Network, 2011, 3, 341-350
doi:10 .4236/wsn.201 1.3100 38 Published Online Oc tobe r 20 11 (htt p://www .Sc iRP.org/jo urna l/wsn)
Copyright © 2011 SciRes. WSN
A Novel Energy Efficient Routing Protocol in Wireless
Sensor Networks
Ali Norouzi1, Faezeh Sadat Babamir2, Abdul Halim Zaim3
1Department of Computer Engineering, Istanbul University/Avcilar, Istanbul, Turkey
2Department of Computer Science, Shahid Beheshti University of Tehran, Tehran, Iran
3Department of Computer Engineering, Istanbul Commerce University/Eminonu, Istanbul, Turkey
E-mail: norouzi@cscrs.itu.edu.tr, babamir@mail.sbu.ac.ir, azaim@iticu.edu.tr
Received September 1, 2011; revised September 21, 2011; accepted September 30, 2011
Abstract
Wireless sensor networks are employed in several applications, including military, medical, environmental
and household. In all these applications, energy usage is the determining factor in the performance of wire-
less sensor networks. Consequently, methods of data routing and transferring to the base station are very
important because the sensor nodes run on battery power and the energy available for sensors is limited. In
this paper we intend to propose a new protocol called Fair Efficient Location-based Gossiping (FELGossip-
ing) to address the problems of Gossiping and its extensions. We show how our approach increases the net-
work energy and as a result maximizes the network life time in comparison with its counterparts. In addition,
we show that the energy is balanced (fairly) between nodes. Saving the nodes energy leads to an increase in
the node life in the network, in comparison with the other protocols. Furthermore, the protocol reduces
propagation delay and loss of packets.
Keywords: Wireless Sensor Network, Gossiping, Routing Algorithm, Fair Energy Consumption
1. Introduction
Wireless Sensor Networks consist of tiny sensor nodes
that, in turn, consist of sensors (temperature, light, humi-
dity, radiation, etc.), microprocessor, memory, transcei-
vers, and power supply. In order to realize the existing
and potential application for WSNs, advanced and
extremely efficient communication protocols are required
[1]. WSNs are application-specific, where the design
requirements of WSNs change according to the applica-
tion. Hence, routing protocol requirements are changed
from one application to another. For instance, the re-
quirements of routing protocols designed for environ-
mental applications are different in many aspects from
those designed for military or health applications [2-4].
However, routing protocols for all Wireless Sensor net-
works, regardless of the application, must try to maxi-
mize the network life time and minimize the overall
energy consumption in the network. Network lifetime is
a critical concern in the design of WSNs. In many
applications, replacing or recharging sensors is some-
times impossible [5]. Therefore, many protocols have
been proposed to increase network lifetime. It is difficult
to analyze network lifetime because it depends on many
factors, like network architecture and protocols, data
collection initiation, lifetime definition, channel charac-
teristics, and the energy consumption model [6]. For all
routing protocols, energy consumption during communi-
cation is a major energy depletion parameter; the number
of transmissions must be reduced as much as possible to
achieve extended battery life. For these reasons, the
energy consumption parameter is a top priority [7].
In this article, we propose a new routing protocol
based on Gossiping called Fair Efficient Location-based
Gossiping (FELGossiping) to improve the problems of
Gossiping and its extensions. FELGossiping consists of
three phases: Initialization, Information Gathering and
Routing. In the first phase, each node generates the
gradient to the sink. In the second phase, the FEL
Gossiping sends a request message to the other nodes to
receive the information of other members or neighboring
nodes. Once the hop count and the remainder energy of
the member nodes are known, FELGossiping chooses
two nodes in the third phase. The nodes are chosen near
to the base station, according to the hop count of the
selected nodes with the sink node, in order to deliver the
A. NOROUZI ET AL.
342
packet to the sink. After selecting two nodes, the pro-
tocol only chooses one of the two nodes to send the
packet. The node with more residual energy is selected,
and the message is sent to the selected node to broadcast
the packet to the base station. Finally, we present some
optimal strategies and through simulation results show
that the optimal routing strategies provide a significant
benefit.
This paper is organized as follows. In Section 2 we
give a brief description of the energy efficient routing
protocols, especially GOSSIPING. Section 3 presents
our proposed algorithm and its details. Section 4 shows
the methodology we followed to perform the simulations
and the results are also provided in this Section. Finally,
Section 5 presents our conclusions and suggestions for
future projects.
2. Literature
During previous research, many differences have been
observed, generally, between flat and hierarchical rout-
ing protocols and, exactly, between these researched
routing protocols [8,9]. In this paper, we choose Gossip-
ing as a target protocol to conduct our research and some
extensions. Firstly a technical glimpse on Gossiping:
Gossiping is a data-relay protocol, based on a Flooding
protocol, and does not need routing tables or topology
maintenance [10]. It was produced as an enhancement
for Flooding and to overcome the drawbacks of Flooding,
i.e., implosion [11]. In Flooding, a node broadcasts the
data to all of its neighbors even if the receiving node has
just received the same data from another node. The
broadcasting will continue until the data is received by
the destination [1,12]. However, in Gossiping, a node
randomly chooses one of its neighbors to forward the
packet to, once the selected neighbor node receives the
packet it, in turn, chooses another random neighbor and
forwards the packet to it. This process will continue until
the destination or number of hops has been exceeded. As
a result, only the selected nodes/neighbors will forward
the received packet to the sink [13]. Unlike Flooding,
Gossiping operates well in a one-to-one communication
scenario but it does not in a one-to-many. Packet for-
warding mechanisms for both Flooding and Gossiping
are shown in Figure 1 [4,14].
Gossiping consumes less energy than Flooding. How-
ever, it suffers from latency; information propagates
slowly, one node at each step. Despite the simplicity and
inefficiency of Flooding and Gossiping, they could be
used for specific functions, for example, during deploy-
ment phases and network initialization [11,12]. The power
consumed by Gossiping, is approximately equal to O
(KL).
Figure 1. Forwarding mechanisms of Flooding and Gossip-
ing.
K: number of nodes that forward the packet.
L: number of hops before the forwarding stops.
The most remarkable feature of Gossiping is the abil-
ity to control the power consumption by appropriately
selecting K and L [15].
After a technical review of Gossiping protocols we
can determine these disadvantages:
The next hop neighbor is randomly chosen, this
means it may include the source itself.
The packet will travel through these selected neigh-
bors until it reaches the sink or exceeds the number of
hops.
It suffers fr om packet lo ss.
The most sig nificant disadvantage of Gossiping is t hat
it suffers from latency caused b y data pro p a gation.
Finally, in order to enhance the Gossiping protocol,
many protocols have been produced as an expending to it.
For example FLossiping [16], SGDF [17], LGossiping
[18] and ELossiping [19].
FLossiping Protocol: combines the approaches of both
the Flooding and Gossiping routing protocols. When a
node has a packet to send, it chooses a threshold and
saves it in the packet header, it then randomly selects a
neighbor to send the packet in Gossiping mode, while the
other neighbor nodes listen to this packet and generate a
random number. The neighbors whose randomly gene-
rated numbers are smaller than the threshold will broad-
Copyright © 2011 SciRes. WSN
A. NOROUZI ET AL.343
cast the packet in Flooding mode. As a result, FLOS-
SIPING improves the packet overhead in Flooding and
the delay issue in Gossiping [1 6 ].
SGDF Protocol: Single Gossiping with Directional
Flooding routing protocol is divided into two phases;
Net work T opo logy Ini tial izat io n and Ro uting Sc he me. I n
the first phase, each node generates a gradient (showing
the number of hops to the sink). In the second phase, in
order to deliver the packet, SGDF uses single gossiping
and directional flooding routing schemes. As a result (see
Figure 2), SGDF achieves a high packet delivery ratio,
low message complexity, and short packet delay [17].
However, the disagreeable side effect of this protocol is
that the amount of packets becomes larger during packet
delivery due to directional flooding.
LGossipi ng Protocol: in Location based Gossiping
protocol, when a node has an event to send it randomly
chooses a neighboring node within its transmission ra-
dius. Once the neighbor node receives this event, it in
turn randomly chooses another node within its transmis-
sion radius and sends it. This process will continue until
the sink is reached. As a result, the delay problem has
been solved to some extent. Figure 3 shows the main
objective of LGossiping [18].
Although in this protocol the delay problem has been
solved to some extent, there is still the problem of many
events not reaching the main station. Moreover, this pro-
tocol uses GPS to determine the location of each node.
Hence, additional hardware is required, which means
extra money.
ELGossiping Protocol: in the Energy location base
Gossiping protocol, when a node detects an event and
wants to send information, it selects a neighboring node
within its transmission radius and the lowest distance to
the base station/sink. Once the neighboring node receives
the event, it will in turn select another neighboring node
within its transmission radius and the lowest distance to
the sink. The event will travel in the same manner to the
sink. As a result, the problem of latency and the situation
Figure 2. Routing scenario in SGDF.
Figure 3. Schematic of data routing in LGossiping.
of non-reaching packets have to some extent been solved
[19]. In Figure 4 you can see details of this protocol.
Two important metrics have been exploited in this
protocol; Energy and distance to the base station, and in
this way when a node detects an event within its trans-
mission range it sends the data to a neighboring node that
has a shorter distance to the sink.
Although in this protocol and its extensions some of
the problems have been solved, there is still a problem of
many packets not reaching the main station.
3. New Algorithm
With some changes to the Gossiping protocol we can
decrease the energy consumption and also increase the
network lifetime. Therefore, in order to resolve the dra w-
backs of the Gossiping protocol, we have proposed a
new protocol as an extension for Gossiping. In this pro-
tocol we have increased the network lifetime by selec ting
a node with a maximum residual energy and lower dis-
tance to the sink. We have also achieved a high packet
delivery ratio and reduced the delay in delivering the
packet.
The new algorithm consists of three phases: Network
Initialization Phase, Information Gathering Phase and
Routing Phase. In this section we explain three parts of
the algorithm as follows.
Network Initialization Phase
The network initialization phase starts after the sensor
nodes are randomly distributed in the application area. In
the beginning, the base station broadcasts a “HELLO”
message to its neighbors. The HELLO message contains:
the Base Station Address (fixed) and Hop Count (vari-
able). The hop count is used to setup the gradient to the
base station, which means it shows the node distance to
Copyright © 2011 SciRes. WSN
A. NOROUZI ET AL.
Copyright © 2011 SciRes. WSN
344
will discard the message. This case occurs due the mes-
sage ha s bro adcast ed previ ously thro ugh diffe rent r outes.
As a result, the gradient will keep the best route. Finally,
the process will continue until all the sensors receive the
HELLO message, at that time the network initialization
phase will be completed. Now each node through the
gradient knows its distance to the base station. Figure 6
summarizes the setup phase in a simple flowchart. Fig-
ure 5 shows how the HELLO message is broadcast
through the network. In Figure 5(a) 1-hop neighbor to
the Base Station will store the hop count in its memory
and start constructing the gradient. Each node increments
the message hop count by 1 and broadcasts the message
with the new hop count (at this time is has become 2). In
Figure 5(b) we can see that the HELLO message has
reached 1-hop neighbors; two of them become broad-
casting nodes which in turn continue broadcasting the
message. Then the neighbors of the broadcasting node
receive the message and compare the hop count of the
message with their own. If it is smaller than its hop count,
they will discard the message as shown in the Figure 5.
The process will continue until completing the initialize-
tion phase as shown in Figure 5(c). The network ini-
tialization phase is as shown in the following flow chat
in Figure 6.
Figure 4. Routing in ELGossiping.
the base station. After broadcasting the HELLO massage,
all 1-hop neighbor s will receive this messa ge and get the
base station address and the hop count. Each node saves
the hop count in its memory and increases the hop count
by 1. The new hop count is then replaced with the old
one. After each node has received the HELLO message it
will continue to broadcast this message to farther nodes.
As shown in Figure 5, at each stage the hop count will
be incremented by 1.
Information Gathering Phase
After detecting the event the source node will draw a
transmission radius of 40 m to deal with the nearby
nodes.
All sensors have GPS and can move to any position
within their mobility range. The source node then gene-
rates the request message to acquire the information fro m
the neighboring nodes in its transmission radius. The
request message is contained in the hop count of the
neighboring node to the sink, or the distance of nodes
from the sink and the residual energy of all nodes. After
that, the nodes that received the request message send
When a node receives a HELLO message it will check
whether it already has a gradient. If it has a gradient, it
will compare it to the hop count of its own message and
will replace its hop count with the message’s hop count
if the latter is smaller, and will add 1 to the hop count
prior to broadcasting it. However, if its hop count is
smaller than or equal to the hop count of the message, it
Figure 5. Network initialization phase.
A. NOROUZI ET AL.345
Fig ure 6. Flowchart of the network initi alization phase.
their information to the source node or to the node that
detected the event. The information gathering phase is as
shown in the following flow chat in Figure 7.
Routing Phase
After the network request phase finishes, the routing
phase will start. Following we state some assu mptio ns:
- At the start all nodes are full of energy and have the
same amount of energy.
- Each node knows its remaining energy at any stage of
its life.
- Each node has a transmission radius of 40 m.
After this, the receiving nighbor replies to this re- e
Copyright © 2011 SciRes. WSN
A. NOROUZI ET AL.
346
Figure 7. Flowc har t of information gathering phase and rout ing phase.
quest by using its residual energy and hop count. Next,
the source chooses within its transmission radius two
neighboring nodes that have the minimum hop count
towards the sink. Figure 9 shows the selection of valid
nodes in a defined radius. The hop count of one node
will be a unit lower than the other node, or equivalent.
After choosing two nodes near the sink, we compare
between these two nodes according to the residual
energy of the two nodes. Figure 10 shows this op eration
clearly among two selected nodes, we select the nodes
that had the most residual energy and we ignore the
maximum hop count of the two nodes. If two nodes have
the same residual energy we take the nodes that have a
lower hop count to the sink. After that the source node
Copyright © 2011 SciRes. WSN
A. NOROUZI ET AL.347
sends the packet to the selected node. Upon receiving the
message the node repeats the information gathering
phase and routing phase processing to transmit the mes-
sage to another node. Figure 11 shows the routing ope-
ration. The process continues until the message reaches
the sink or the TTL is finished. The sent packet is
including message and packet header. The header packet
is shown in Figure 8. T he he ader consis ts o f six fie ld s as
shown below:
Figure 8 . Header packet data.
Figure 9. Source node outlining its transmission radius.
Figure 10 . Source node within its trans mission radius selec-
ting tw o nodes.
Figure 11. Routing phase.
Source: Source address of the source node (event
node).
Destination: Address of the sink.
Present: Address of the current node.
Next: Address of the next hop.
Hop: Hop count from the current node to the base sta-
tion (number).
TTL: Time to live of the packet. The packet will be
dropped if this field becomes 0.
As soon as the next hop is selected its address will be
written in the next field of the header. The source will
write its address in the source field and the present field,
just as information for the node that detected the event.
After which the source field (address of the event de-
tected node) will be fixed, but the present field will
change according to the present node. The hop count of
the next node will be written in the hop field of the
header. Finally, will be put a value in the TTL1 field,
taking into account that the source will not find a
neighbor with hop count lower than its hop count, in this
case it selects a neighbor with a hop count equal to its
own. Each node receiving the packet will subtract 1 fro m
the TTL field and will progress the packet towards the
base station. The packet relaying will continue until the
Base Station receives the packet, or the TTL field be-
comes 0.
When the base station receives the packet, it will
broadcast a declaration message of successful reception
to inform other nodes trying to send the same packet
through another route to drop the packet. This declara-
tion message prevents the same message reaching the
1Time to Live of the packets.
Copyright © 2011 SciRes. WSN
A. NOROUZI ET AL.
Copyright © 2011 SciRes. WSN
348
base station twice, reduces the message overhead and the
energy consumed by the other nodes during packet
transmission thro ugh other rout e.
4. Simulation and Evaluation
We assu me some the para mete rs to implement the simu-
lation results:
Radius: the radius of coverage of the sensors is 40 m.
Any sensor that detects the event draws the transmission
radius to limit the number of nodes in its transmission
range.
Residual Energy: shows the amount of energy re-
maining. We assume that in the beginning the energy of
all nodes is sa me.
Location of sensor: all sensors have GP S or other lo-
cation devices and can move to any position (with known
coordinates) within their mobility range. It can be shown
that the proposed FELGossiping routing protocol per-
forms better when compared to the other routing pro-
tocols. We investigated the performance of the proposed
protocol by comparing packet loss, delay, live node and
total energy saving per round.
Packet loss: as seen in Figure 12, packet loss is at a
minimum in FELGossiping when compared with Gos-
siping, LGossiping and ELGossiping. However, the
packet loss increased after the 500th iteration.
Live nodes: after the 100th iteration in Gossiping al-
most all nodes die. However, as shown in Figure 13, in
our proposed protocol the nodes die after approximately
750 iterations by balancing and using the energy in a fair
manner.
Delay: in contrast to other protocols that randomly se-
lect the next hop for the packet, in our protocol the near-
est node to the base station is selected as the next hop.
Moreover, we have used GPS to find the base sta-
Figure 12. Packet loss.
Figure 13. Amount of live nodes.
A. NOROUZI ET AL.
Copyright © 2011 SciRes. WSN
349
tion address in an efficient manner. Therefore the delay
in our protocol is the smallest among the protocols com-
pared. We can see this clearly from the simulation results
in Figure 14. Until the 450th iteration the delay is fixed
approximately at 2msec.
Energy Consumption: in our proposed protocol the
relay nodes are not selected blindly (without knowing
their residual energy) as is done in the other routing pro-
tocols in this comparison. Moreover, energy reduction
for each node occurs for every transmission or reception
made. Hence, the probability of choosing the same node
as the next hop is reduced. Thereby, the energy has been
balanced and fairly used. All this leads to saving energy
and hence prolonging the overall network lifetime com-
pared to the other protocols, as seen in Figure 15.
5. Conclusions and Proposals for Future
Work
Wireless Sensor Networks are powered by the limited
capacity of batteries. Due to the power management
activities of these sensor nodes, the network topology
changes dynamically. These essential properties pose
additional challenges to co mmunication proto cols. In this
article we studied the operation of a Gossiping routing
protocol with safe energy consumption, and discussed
the factors of energy optimization. By carefully attending
to the Gossiping protocol we find that by altering the
ways i n which we choose the next ho p, the net work life -
time can be extended.
As a result, in our proposed protocol; we firstly have
Figure 14. Packet delay.
Figure 15. Energy consumption.
A. NOROUZI ET AL.
Copyright © 2011 SciRes. WSN
350
extended the network lifetime through fair use of the
energy by selecting nodes with the maximum residual
energy and lowest distance to the sink. Secondly, we
have achieved a high packet delivery ratio (number of
non-reaching nodes has been reduced) and reduced the
delay in delivering the packet. Thirdly, we have reduced
the message overheads and the energy consumed by the
nodes that have already tried to send the data to the base
station by sending an acknowledgement message of the
successful reception of the packet. At this time we would
like to introduce the Green Network Project as a future
project. In Green Wireless Networks” we propose a new
routing protocol that optimizes energy consumption and
bandwidth. Using less energy in routing protocols reduce
nature pests.
6. References
[1] I. F. Akyildiz and M. C. Vuran, “Wireless Sensor Net-
wor ks ,” 1st Edition, John Wiley & Sons, Ltd, Chichester,
2010.
doi:10.1002/9780470515181
[2] Á. Léd ec zi , A. N ád as , P . V öl gye s i, G. B al o gh , B. Ku sy, J.
Sallai, G. Pap, S. Dóra, K. Molnár, M. Maróti and G
Simon, “Countersniper System for Urban Warfare,” ACM
Transactions on Sensor Networks, Vol. 1, No. 2, 2005, pp.
153-177. doi:10.1145/1105688.1105689
[3] T. He, S. Krishnamurthy, J. A. St ankovic, T. Abdel zaher,
L. Luo, R. Stoleru, T. Yan, L. Gu, G. Zhou, J. Hui and B.
Krogh, “VigilNet: An Integrated Sensor Network System
for Energy-Efficient Surveillance,” ACM Transactions on
Sensor Networks, Vol. 2, N o. 1, 2 0 06 , pp. 1-38.
doi:10.1145/1138127.1138128
[4] R. Verdone, D. Dard ari, G. Mazzini and A. Cont , (2007).
“Wireless Sensor and Actuator Networks Technology,
Analysis and Design,” 1st Edition, Elsevier, London,
2007.
[5] S. Bandyopadhyay and E. Coyle, “An Energy Efficient
Hierarchical Clustering Algorithm for Wireless Sensor
Networks,” Proceedings of the 22nd Annual Joint Confer-
ence of the IEEE Computer and Communications Societies
(INFOCOM 2003), San Francisco, 30 March-3 April 2003,
pp. 1713-1 723.
[6] Y. Chen and Q. Zhao, “On the Lifetime of Wireless Sen-
sor Networks,” IEEE Communications Letters, Vol. 9, No.
11, 2005, pp. 97 6- 9 78 .
do i:10.1109/LCOM M.2005.11010
[7] A. Norouzi and A. Sertbas, “An Integrated Survey in
Efficient Energy Management for WSN Using Architec-
ture Approach,” International Journal of Advanced Net-
working and Applications, Vol. 3, No. 1, 2011, pp. 968-
977.
[8] A. Kanavalli, D. Sserubiri, P. D. Shenoy, K. R. Venugopal
and L. M. Patnaik, “A Flat Routing Protocol for Sensor
Networks,” Proceeding of International Conference on
Methods and Models in Computer Science, Delhi, 14-15
December 2009, pp. 1-5 .
[9] S. Banerjee and S. Khuller, “A Clustering Scheme for
Hierarchical Control in Multi-hop Wireless Networks,”
Proceedings of 20th Joint Conference of the IEEE Com-
puter and Communications Societies (INFOCOM’ 01),
Anchorage, 22-26 April 2001, pp. 1028-1037.
[10] J. M. Rabaey, M. J. Ammer, J. L. da Silva, D. Patel and S.
Roundy, “PicoRadio Supports Ad Hoc Ultra-Low Power
Wireless Networking,” IEEE Computer, Vol. 33, No. 7,
2000, pp. 42- 4 8.
[11] W. Rabiner Heinzelman, J. Kulik and H. Balakrishnan,
“Adaptive Protocols for Information Dissemination in
Wireless Sensor Networks,” Proceedings of the Fifth
Annual ACM/IEEE International Conference on Mobile
Computing and Networking (MobiCom ’99), Seattle,
Was hing ton, 15-20 Aug us t 199 9, pp. 17 4-185.
[12] S. M. Hedetniemi, S. T. Hedetniemi and A. L. Liestman,
“A Survey of Gossiping and Broadcasting in Communica-
tion Networks,” Networks, Vol. 18, No. 4, 1988, pp. 319-
349. doi:10.1002/net.3230180406
[13] K. Akkaya and M. Younis, “A Survey of Routing Proto-
cols in Wireless Sensor Networks,” Elsevier Ad Hoc
Network Journal, Vol. 3, N o. 3, 2 005, pp. 325- 3 49.
do i:10.1016/j.adhoc.2003.09 .010
[14] K. Sohraby, D. Minoli and T. Znati, “Wireless Sensor
Networks Technology, Protocols, and Applications,”
John Wiley & Sons, Inc., Hoboken, 2007.
doi:10.1002/047011276X
[15] W. Heinzelman, A. Chandrakasan and H. Balakrishnan,
“An Application-Specific Protocol Architecture for Wire-
less Microsensor Networks,” IEEE Transactions on Wire-
less C omm unications , Vol. 1, No. 4, 200 2, p p. 66 0- 67 0.
doi:10.1109/TWC.2002.804190
[16] Y. C. Zhang and L. Cheng, “Flossiping: A New Routing
Protocol for Wireless Sensor Networks,” Proceedings of
the 2004 IEEE International Conference on Networking,
Sensing & Control Taipei, Taipei, March 21-23 2004, pp.
1218-1223.
[17] W. Yen, C.-W. Chen and C.-H. Yang, “Single Gossiping
with Directional Flooding Routing Protocol in Wireless
Sensor Networks,” 3rd IEEE Conference on Industrial
Electronics and Applications, Taipei, 3-5 June 2008, pp.
1604-1609.
[18] S. Kheiri, M. B. Ghaznavi, G. M. Rafiee and B. Seyfe,
“An Improved Gossiping Data Distribution Technique
with Emphasis on Reliability and Resource Constraints,”
IEEE 2009 International Conference on Communications
and Mobile Computing, Singapore, 6-8 January 2009, pp.
1604-1609.
doi:10.1109/ICIEA.2008.4582790
[19] A. Norouzi, M. Dabbaghian, A. Hatamizadeh and B. B.
Ustundag, “An Improved ELGossiping Data Distribution
Technique with Emphasis on Reliability and Resource
Constraints in Wireless Sensor Network,” 2010 Interna-
tional Conference on Electronic Computer Technology
(ICECT), Kuala Lumpur, 7-10 May 2010, pp. 179-183.
doi:10.1109/ICECTECH.2010.5479964