Wireless Sensor Network, 2011, 3, 209-214
doi:10.4236/wsn.2011.36024 Published Online June 2011 (htt p://www.SciRP.org/journa l / wsn)
Cr oss Layer Design for Cooperative Transmission in
Wireless Sensor Networks
Kanojia Sindhuben Babulal, Rajiv Ranjan Tewari
Departm e nt of E lectronics and Communication, J.K Institute of Applied Physics and Technology,
University of Allahabad, Allahabad, India
Received April 19, 2011; revised June 1, 2011; accepted June 9, 2011
Abstract
Several protocols and schemes have been proposed to reduce energy consumption in Wireless Sensor Net-
works (WSNs). In this pap er we employ farcoopt, a cross lay er design approach with the concep t of cooper-
ation among the nodes with best farthest neighbor scheme to increase the Quality of Service (QoS), reduce
energy consumptio n, i ncreases p erfo r man ce an d en d-to-end throughput. We present cooperative transmissio n
to connect previously disconnect parts of a network thus overcoming the separation problem of multi-hop
network. We show that this approach improves connectivity over 50% compared to multi-hop approaches
and reduces the number of nodes necessary to provide full coverage o f an area up to 35%. Simulation results
show that on increase of data rates i.e. pack et the net work life t ime increa ses in farcoopt as co mpared to t ra-
ditional multi hop approach. The result of this analysis is presented in this work.
Keywords: Cooperative Network, Cross Layer Design, Wireless Sensor Networks, Energy Saving,
Communication Protocol, Rout ing
1. Introduction
During last few years processor power consumption has
increased by over 200% every four years, while battery
energy density has increased by modest 25% [1]. In
Wireless Sensor Network (WSNs), sensor nodes are
deployed in a hostile environment which comes up with
the problem charging the battery supported equipments.
In fact it has been observed that power is and will always
remain a problem to be solved. To deal with this problem,
there have been increasing interests in design for wire-
less networks that rely on interactions between various
layers of the protocol stack. This approach called cross
layer design has been widely recognized as a promising
solution for various problems in wireless networks.
However, such a design principle across different layers
usually involves high complexity [2]. Cross layer design
is currently one of the most active research areas in
computer networks. Cross layering means allowing
communication of layers with any other layer [3]. Cross
Layering was thought of to address QoS (Quality of Ser-
vice), Poor performance, wireless links, mobility, packet
loss, delay problems observed in wireless networks. The
traditional method to forward data in sensor network is
multi hop routing. Many MAC protocols and routing
algorithm for WSN are developed like PAMAS [4],
SMAC [5]. Many network layer algorithms such as MTE
[6], DSR [7], and AODV [8] try to decrease the total
energy consumption by multi hop routing. Nodes try to
find a route between source and a destination to forward
the information. For communication at least one route
must exist. If two nodes are not able to communicate to
each other because they are too far away to send/receive
RF signals the network is segmented [9].
2. Related Works and Background
In research paper [10] an energy efficient routing algo-
rithm is proposed which is based on the concept of
switching transmission power level based on the volume
of data. It proposes a back off mechanism by switching
OFF unintended receivers based on the power of the re-
ceived radio signal. In [11], the cross layer interaction is
shown with MAC and the routing layer. The cross layer
design is implemented in this paper by showing the inte-
raction of 802.11 MAC protocol and Dynamic Source
Routing (DSR) protocol. This design was able to reduce
the routing overheads by decrementing the route man-

K. S. BABULAL ET AL.
210
agement process performed by the DSR protocol. In pa-
per [12], an energy optimization protocol based on cross
layer wireless sensor networks named as EOA was pro-
posed. For this physical, MAC and the routing layer were
considered. A feedback algorithm is proposed which
computes proper transmission power level between
nodes was computed. Then the EOA routing protocol can
make use of the transmission power as a metric by
choosing route with optimal power consumption to for-
ward packets. In research paper [13] an efficient flooding
scheme which reduce the re-transmission redundancy are
introduced. In this nod e can decide if it should r etransmit
the broadcast packet or not. In probabilistic flooding,
node re transmits the broadcast packet according to
probability. In [1 4] energy efficient cross layer routing is
obtained by combining the MAC functionality with the
routing decision at the network layer. By this the over-
head is reduced to some level. The protocol used is based
on geographic routing and the end-to-end decision is
made by the source node and the dest i na tion node.
In [15] a cross layer design frame work called Coop-
Geo, which performs the greedy forwarding nodes that
wins the contention to forward the message, cooperative
relaying scheme with single relay selection mechanism
where the source node and relay node jointly transmit
data through wireless channel is used. CoopGeo gains
physical layer performance in terms of reliability. In [16,
17] authors understand cooperative transmission in the
sense that several sensor nodes transmit symbols simul-
taneously to achieve the power gain. In [17] the broad-
cast coverage of a system using cooperative transmission
is analyzed. It is based on a continuum approach model-
ing the nodes as a homogenous density of possible
transmits power. This simplifies the modeling and leads
to closed-form solution and formulations. In [18] there is
also a small section on the coverage achievable with co-
operative transmission.
Our work is based on the maximum intermediate node
concept [18-21]. In these researchers, the authors’ pro-
posed novel concept for routing i.e. instead of using the
traditional method for routing they introduced a new
concept of sending the data to the maximum distance
intermediate node. In [19] they have also applied a dy-
namic re-transmission scheme also. And an important
thing to note is that in those they have considered the
networks that are static or less mobile. In our paper we
have talked about the network that is mobile in nature,
and applied cooperative transmission scheme which
helps to increase the QoS and reliability of the system.
The rest of the paper is organized as below: Section 3
discusses briefly about the cooperative transmission
scheme. Section 4 presents the detail of the network
model that is used in WSN and the detail description of
FARCOOPT scheme. Section 5 shows the simulation of
the scheme. Section 6 discusses the result of the simula-
tions.
3. Network Model & Cooperative
Transmission Theme
Cooperative transmission theme is applied to the net-
works that are not able to communicate with each other.
It is in ideal means applied to tackle the bad connectivity.
Connection break can be from any of the reasons like
random installation process, changes in the environment,
wear-out or due to the mobility of the networks. With
cooperative transmission, a group of nodes can combine
its emission power and achieve a higher emission power
as a whole. To do so, cooperatively transmitting nodes
emit identical symbols synchronously to superimpose the
emitted waves on the physical medium. The destination
receives the sum of waves and thus a higher total power.
The more nodes cooperatively transmit the higher will
the power on the physical medium be. With the higher
power, the nodes can reach destinations that are very far
away.
In this paper, we address the concern of cross layer
design with wireless medium of physical layer and MAC
sub layer being passed to the network layer and the in-
formation of network being transmitted to lower layers.
We aim to improve interactions of physical, MAC, and
network layers. Information about the physical channel
condition is transmitted from physical layer to network
layer. Data rate and power information are transmitted
from network layer down to the physical interface.
We model the dynamic wireless model as a set of N
nodes distributed into two dimensional planes. Wireless
network is represented as a graph G(V, E), where V =
{v1, v2, v3,
⋅⋅⋅
, vn} is a finite set of nodes and E = {e1,
e2, e3,
⋅⋅⋅
, en} a finite set of links, the sink and nodes
are randomly deployed in the areas. As wireless net-
works do not hav e fixed infrastructur e, sensor nodes col-
Figure 1. Increased emission range by summation of radio
range.

K. S. BABULAL ET AL.
211
laborate to work together by wireless channel. Every
node is aware of its location and also the location of the
neighbor. The network is also mobile in nature.
The set of nodes source Vsource = {VS1, VS2,
⋅⋅⋅
, VSn}
knows the destination location. In this network Vsource is
sending a set of packets to the destinations d(i) nodes.
Each source node i transmit to its destination node d(i)
which may or may or not belong to S. Let hij denote the
channel gain power between node i and j. all nodes are
assumed to be energy constrained and the total energy of
each node is denoted by Etotal. Note that the energy allo-
cated to a node is the total en ergy utilized in trans mitting
and relaying this node information. All nodes are fully
charged initially and they have same energy. Since for a
particular source destination (s-d) pair, some nodes may
be far away from both source and destination only
neighbor are selected in order to increase power effi-
ciency and avoid err or propagation.
We assume Atx as the nominal transmit power of a
node. We assume that the transmit power is same for all
nodes.
,rx j
Ai←
, is the received power of a signal propagated
from node i to node j. a receive power
,rx j
Pi←
, above
a given threshold Pth will provide sufficient SNR in re-
ceiver to decode the transmission. For cooperating
transmission scheme power gain can be completely ex-
ploited. A group G of nodes all connected to each other
can combine their power and transmits toward a node j:
,,
rx jrx j
iG
PGPi
∈
←= ←
∑
(1 )
In our wireless communication system at physical
layer we consider path loss model. Path loss model is the
difference between the transmitted power and the re-
ceived power. It represents signal l level attenuation
caused by free space propagation, reflection, diffraction
and scattering. Path loss in its simplest form can be de-
scribed as below
( )
10
10 logL ndC= +
(2)
where L is the path loss in decibels, n is the path loss
exponent, d is the distance between the transmitter and
the receiver, usually measured in meters, and C is a con-
stant which accounts for system losses
For radio and antenna consideration path loss can be
calculated as below
( )
10 4
20log
d
L
λ
π
=
(3)
where L is the path loss in decibels, λ is the wavelength
and d is the transmitter-receiver distance in the same
units as the wavelength. For path loss which we model as
a radial fading , 1ij
β
γ . With this the maximum distance
for successful communication is
tx
th th
P
P
β
γ =
. For channel
we assume Rayleigh fading. It models the effect of a
propagation environment on a radio signal and assumes
the magnitude of a signal that has passed through com-
munication channel will randomly fade according to the
Rayleigh distribution. Most mobile communication oc-
curs when there is no direct path between the base station
antenna and the mobile user. The signal reflects off many
objects along the path between the two. This propagation
follows a Rayleigh probability distribution about the
mean signal level:
( ) [ ]
( )
2
22
2
2
exp prob
2
1 exp2
r
R
rr
prr R
R
PR
αα
α

= <



= =−−


R is the signal level, α the value of the peak in the dis-
tribution, with mean 2 α π= and median
M
R=
( )
2ln 21.774
αα
=. The antennas of wireless sensor
nodes are trans-receiver, Omni directional behavior. We
consider that transmission of a node takes a certain
amount of energy Etx. Hybrid cooperative communica-
tions consist of three parts: a single source S, a single
destination D, and set of potential relay nodes B. A pack-
et fails only when there is interference on intended re-
ceiver. Even if the packets collide partially they are con-
sidered to be collided. We assume that throughout the
process there is some mechanism that notifies the sender
of the success and failure of its transmission. We have
assumed initially that the ne twork is mobile in nature and
no fixed infrastructure exit.
4. Farcoopt Scheme
Farcoopt is divided in two major subsections
(4.1) Sendin g the da ta
(4.2) Cooperative Transmission applied in case
(4.2.1) when retransmission occurs
(4.2.2) when link failure occurs
Whole process is explained as below. As we have as-
sumed earlier that our network is mobile so the chances
of link breakages also increases. Let us assume Figure 2.
Case 4.1: want to send the data from source S(1) to des-
tination D(10). First the shortest path from S to D is ob-
tained by using algorithm all pair shortest path. Which is
in this example set of nodes {1, 2, 3, 6, 7, 9, 10}. But
Figure 2. Network consists of 10 nodes where source (1) and
destination (10).

K. S. BABULAL ET AL.
212
instead of using the simple multi hop (node by node)
routing we use the best farthest neighbor concept. Best
farthest neighbor can be assumed as that neighbor who is
active (i.e. free to receive the data), charged, it is in
transmission or radio range. The benefit of best farthest
neighbor is that it reduces the overhead and instead of
sending data to the node next with same energy why not
to send to that node which is far not successfully receive
the data, and to some extend it also help to save energy
as while the nodes are communicating other nodes can
sleep. This decreases the overhead. When any commu-
nication is going on between two nodes other neighbor
nodes do not do any transaction i.e. Case 4.2 when co-
operative transmission takes places they (best farthest
neighbor) need to be active. As it is assumed that trans-
mission of data is with same energy no matter what data
has to be send. Before the sending of the data occurs it is
assumed that procedure of sending the request to check
the status of the best farthest neighbor, and on the receipt
of the acknowledgement that it is free and would receive,
then only data is send. It means that after the confirma-
tion that node is active and free to receive; data is trans-
mitted to that node.
When once the route is discovered S(1) sends data to
node 3 which is S(1) best farthest neighbor. On success-
ful delivery of data node 3 has to send back acknowled-
gement notifying that it has successfully receive the data.
Case 4.2.1 If the acknowledge ment is not received wit h-
in fixed time Tf, it is assumed that packet failure has oc-
curred. Now we need to retransmit the data, so here we
use the cooperative retransmission scheme and send the
data to best farthest neighbor. As we are sending to the
best farthest neighbor the number of relay nodes would
be less which would help in cooperative transmission. To
do so, cooperatively transmitting nodes emit identical
symbol synchronously to superimpose the emitted waves
on the physical medium. And then node 3 notifies with
acknowledgement for successful delivery of data. These
acknowledgement bits are small in size.
Case 4.2.2 Now node 3s best farthest neighbor is node
7, but during communication due to some reason (may
be due to mobility, environment changes etc.) that link
fails and network is cluster they are not able to commu-
nicate with each other, then also the cooperation strategy
is applied to find the nearest link, let us suppose in this
example it is node 8 which would help to join the two
clustered network. The data that is still lying on node 3
has to be forward to node 8. Node 8 also has to notify for
successful receiving of the data. At this stage it should be
made clear that node 8 would not be able to send back
the notification for successful data delivery with the help
of the cooperation means. This step would be costly in
terms of time and energy because we are only sending
the acknowledgement and not the data. But still it would
have to be done. This would increase the (Quality of
Service) QoS an d also the reliability of the network. Im-
portant issue is that we are not proceeding without the
guarantee that the data at that step was successfully re-
ceived. But It can be noted that node 8 was not in the
shortest path which we calculated before the starting of
the process. Again the shortest path would be calculated
and the data would be forwarded as discussed above.
Benefits of best farthest neighbor is that, it reduces the
overhead of control signal, and when same energy we
can transmit at bode situated at farther node , then why to
send it to just the neighbor.
5. Simulations & Results
The set up consist of 250 sensor nodes which are ran-
domly deployed in area 1000 × 1000 m2. We summarize
the configuration setting used as input in our simulator.
Packets are generated by Poisson process. The mobile
rate of nodes is v, we consider that maximum speed of 10
meter per seconds. We assume λe as the average number
of UDP packets generated per second for any source
nodes and packets size of 1024 bytes. We change the
value of λe to show the performance of FARCOOPT.
Network Lifetime is used as metric to evaluate the per-
formance of cross layer design. We define network life-
time as the time taken for 50% of the sensor nodes in a
network to drain up their power. Table 1 gives the de-
tails of the simulation paramete rs .
Figures 3-5 shows the comparison of 3 different
schemes DSR, E2XLRADR, and FARCOOPT method.
All the three schemes are build on different strategy.DSR
uses traditional multi hop routing; E2XLRADR uses the
concept of maximum intermediate node. And FARCOOPT
used the concept of best farthest neighbor with coopera-
tive transmission. We can observe that with the increase
of the node number the lifetime of FARCOOPT increase
in comparison to the other schemes. As the rate increases,
the lifetime of DSR changes very rapidly because of in-
terference, but our FARCOOPT approach can handle
Table 1. Simulation parameters.
Parameters
Value
Simulation area
1000 × 1000 m2
Number of nodes
10, ⋅⋅⋅, 250
Transmission range (M)
250
Bandwidth (Mbps )
2
Max node speed (MIS)
10
Tf (Time fixed) 100
Data item size (KB)
(1 - 10)
Number of data item
1000
Request interval (S)
10
Simulation time (S)
2000
λe1
3packets/sec
λe2 6packets/sec
λe3
9packets/sec

K. S. BABULAL ET AL.
213
Figure 3. Network lifetime of 3 scheme at which node mo-
bile rate is 3 packet/sec.
Figure 4. Network lifetime of 3 scheme at which node mo-
bile rate is 6 packet/sec.
Figure 5. Network lifetime of 3 scheme at which node mo-
bile rate is 9 packet/sec.
Figure 6. Increase in the through put i n FARCOOPT.
Figure 7. On increase of number of packet the quality of
service increases.
with this situation. Figure 6 talks about the throughput in
terms of time. Throughput is not degraded with increase
in time of the process. Throughput is measured in bit/Hz
which is the unit of frequency efficiency. Figure 7 in-
vestigates about the increase in Quality of service even if
the number of packets increases.
6. Conclusions
Cross layering is the best approach for wireless sensor
networks. In traditional layering network various proto-
col layers can only strictly communicate in “layer by
layer” approach. In such conditions, layers are not de-
signed to adapt to the changing environment. This leads
to inefficient use of energy. We believe that cross layer
design should alleviate this problem. We have used the
concept of sending the data to the best farthest neighbor
with the cooperative transmission theme. Cooperative
retransmission has direct impact on the end-to-end delay

K. S. BABULAL ET AL.
214
throughput of WSN. We have used the cooperative re-
transmission when we have to retransmit the data and
when during communication link failure occurs. We
compared our scheme with traditional approaches like
the DSR protocol. By this approach we achieve end to
end packet delivery, less loss of data, reliability of net-
work, QOS, fairness in service, and energy is also saved
to some extent. Several power conservation schemes
have been proposed in the literature for prolonging the
lifetime of the sensor network, either by trying to reduce
the number of retransmission through efficient routing or
by taking the advantages of the sleep mode capabilities
of the sensor node. Network Performance is also in-
creased by FARCOOPT approach.
7. References
[1] K. Lahiri, A. Ragunathan, S. Dey and D. Panigrahi,
“Battery-Driven System Design: A New Frontier in Low
Power Design,” Proceedings of 7th Asia and South Pa-
cific and the 15th International Conference on VLSI De-
sign, Bangalore, 7-11 January 2002, pp. 261-267.
[2] V. Kawadia and P. R. Kumar, “A Cautionary Perspective
on Cross Layer Design,” IEEE Wireless Communications,
Vol. 12, No. 1, 2005, pp. 3-11.
doi:10.1109/MWC.2005.1404568
[3] V. Srivastava and M. Motawani, “Cross Layer Design: A
Survey and the Road Ahead,” IEEE Communications
Magazine, Vol. 43, No. 12, 2006, pp. 112-119.
doi:10.1109/MCOM.2005.1561928
[4] S. Singh and C. S. Raghavendra, “PAMAS: Power Aware
Multi-Access Protocol with Signaling for Ad Hoc Net-
works,” ACM Computer Communication Review, Vol. 28,
No. 3, 1998, pp.5-26. doi:10.1145/293927.293928
[5] W. Ye, J. Heidermann and D. Estrin, “An energy Effi-
cient MAC Protocol for Wireless Sensor Networks,”
Proceedings of IEEE Infocom, USC/Information Scie nces
Institute, New York, June 2002, pp. 1567-1576.
[6] W. Heinzelman, A. Chandrakasan and H. Balakrishnan,
“Energy Efficient Communication Protocol for Wireless
Microsensor Networks,” Proceedings of the Hawaii In-
ternational Conference on System Sciences, Washington
DC, January 2000, pp. 4-7.
http://www.isi.edu/johnh/PAPERS/Ye02a.html
doi:10.1109/HICSS.2000.926982
[7] D. B. Johnson, D. A. Maltz, Y.-C. Hu and J. G. Jetchev,
“The Synamic Source for Mobile Ad Hoc Wireless Net-
works,” Internet Engineering Task Force Internet Draft,
2000.
[8] C. E. Perkins and E. M. Royer, “Ad Hoc on Demand
Distance Vector Routing,” Internet Engineering Task
Force Internet Draft, 2003.
http://www.ietf.org/internetdrifts/draft-ietd-manet-adsr-0
6.txt
[9] R. Gunesekaran and H. Qi, “XLRP: Cross Layer Routing
Protocol for Wireless Sensor Network,” IEEE Communi-
cation Society, Las Vegas, 31 March-3 April 2008, pp.
2135-2140.
http://www.ietf.org/internetdrifts/draft-ietf-manet-aodv-1
3.txt
[10] N. Chilamkurti, S. Zeadally, A. Vasilakos, et al., “Cross
Layer Support for Energy Efficient Routing in Wireless
Sensor Networks,” Journal of Sensors, Vol. 2009, 2009,
pp. 1-9.
[11] Y. B. Bai, S. J. Liu, et al., “An En ergy Optimiza tion Proto-
col Based on Cross Layer for Wireless Sensor Networks,”
Journal of Communication, Vol. 3, No. 6, 2008, pp. 27-34.
doi:10.1155/2009/134165
[12] S. Ni, Y. Tseng, Y. Chen and J. Sheu, “The Broadcast
Storm Problem in Mobile Ad Hoc Network,” Wireless
Netwo rk, Vol. 8, No. 2-3, 2002, pp. 153-167.
[13] P. Skraba, H. K. Aghajan and A. Bahai, “Cross Layer
Optimization for High Density Sensor Network: Distri-
buted Passive Routing Decisions, Ad Hoc, Mobile, and
Wireless Networks,” Proceedings of the Third Interna-
tional Conference, ADHOC-NOW, Vancouver, July 2004,
pp. 266-279.
[14] T. Aguilar, et al., “A Cross Layer Design Based on Geo-
graphic Information for Cooperative Wireless Networks,”
Proceedings of the IEEE 71st Vehicular Technology
Conference, Taiwan, 16-19 May 2010, pp. 1-5.
[15] A. Scaglione and Y. W. Hong, “Opportunistic Large Ar-
rays: Cooperative Transmission in Wireless Multi Hop
Ad Hoc to Reach Far Distances,” IEEE Transaction on
Signal Processing, Vol. 51, No. 8, 2003, pp.2082-2092.
[16] A. Scaglione and Y. W. Hong, “Opportunistic Large Ar-
rays: Cooperative Transmission in Wireless Multi Hop
Ad Hoc to Reach Far Distances,” IEEE Transaction on
Signal Processing, Vol. 51, No. 8, 2003, pp.16-20.
doi:10.1109/TSP.2003.814519
doi:10.1109/TSP.2003.814519
[17] K. S. Babulal and R. R. Tewari, “E2XLRADR (Energy
Efficient Cross Layer Routing Algorithm with Dynamic
Retransmission for Wireless Sensor Network”, Interna-
tional Journal of Wireless & Mobile Network, Vol 2, No.
3, August 2010, pp.166-177. http://www.airccse.org
[18] K. S. Babulal and R. R. Tewari, “Cross Layer Energy
Efficient Routing (XLE2R) for Prolonging Lifetime of
Wireless Sensor Networks,” Procee di ng s of IEEE Interna-
tional Conference on Computer and Communication Tech-
nology, Allahabad, 17-19 September, 2010, pp. 167-177.
[19] K. S. Babulal and R. R. Tewari, “Routing Algorithm: An
Energy Efficient Sleep/Awake Cross Layer for Wireless
Sensor Networks”, Advances in Wireless and Mobile Com-
munications, Vol. 3, No. 2, 2010, pp. 139–146.
http://www.ripublication.com/awmcv3n2.htm
[20] K. S. Babulal, T. R. Ranjan, “Cross Layer RouteDisco-
very Using Farthest Intermediate Link Distance Reliabil-
ity Cost Based for MANETS,” National Academy of Let-
ters, Vol. 34, No. 5-6, 2011.
[21] K. Albert, B. Michael, D. Christian and R. Till, “Increas-
ing Connectivity in Wireless Sensor Network Using Co-
operative Transmission,” The 3rd International Confe-
rence on Networked Sensing Systems (INSS), Chicago, 31
May-2 June 2006.