Download iii. estimation of network lifetime using adaptive duty cycle

International Journal of Advanced Computer Engineering and Communication Technology (IJACECT)
________________________________________________________________________________________________
Comparative Study of Duty Cycle and Network Coding based Analysis of
Lifetime for Wireless Sensor Networks
1
Shashikala A, 2Manoj P B
PG Student1 , Assistant Professor2
Email: [email protected], [email protected]
Abstract- Wireless Sensor Network (WSN) consists of
autonomous sensor nodes which are equipped with limited
battery and requires energy management for enhanced
Network Lifetime. The area around the sink forms a bottleneck
zone due to heavy traffic-flow will be in demand of more power
which limits the network lifetime. This work attempts to
enhance the energy efficiency of the bottleneck zone which
leads to overall improvement of the network lifetime by
considering an adaptive duty cycled WSN. In this paper, an
efficient communication paradigm has been adopted in the
bottleneck zone by combining adaptive duty cycle and network
coding. Comparative studies carried out to estimate the upper
bounds of the network lifetime by considering (i) Adaptive
duty cycle, (ii) Network coding and (iii) Combinations of
adaptive duty cycle and network coding.
The node which adapts the sleep time to the traffic changes
dynamically by considering the queue length at a
predetermined value. Network coding is not simply relaying
the packets of information they receive, the sensor nodes of a
network take several packets and combine them together for
transmission and applied in bottleneck zone. By applying the
above techniques the overall lifetime of the node will eventually
increases. Energy efficiency of the bottleneck zone increases
because more volume of data will be transmitted to the sink
with the same number of transmissions. A theoretical analysis
and simulation results have been carried out to show the
efficacy of the proposed approach in terms of network lifetime.
Energy is a scarce resource in wireless sensor networks and
conservation of energy has been the subject of extensive
research. While a variety of solutions have been proposed,
duty cycling and network coding have proven to be two of
the most successful techniques in this field. In most
applications of sensor networks, the sensing information is
transported only to a sink node, where information that is
helpful for users conveyed to users of the applications. The
sensor nodes near a
sink node suffer from the heavy traffic load imposed on them
and their energy is depleted strongly. This phenomenon is
called the energy hole problem.
In a typical WSN, the network traffic converges at the Sink
node S (Fig. 1). There is a large amount of data flow near the
Sink. The area near the Sink is known as the bottleneck zone.
Heavy traffic load imposes on the sensor nodes near the Sink
node. The nodes in the bottleneck zone vacate their energy
very quickly, referred as energy hole problem in WSN.
Failure of such nodes inside the bottleneck zone leads to
wastage of network energy, bandwidth and reduction of
network reliability. The bottleneck zone needs special
attention for reduction of traffic which improves the network
lifetime of the whole WSN.
Index terms- Duty cycle, network coding, network lifetime,
Wireless sensor networks.
I. INTRODUCTION
Wireless Sensor Networks (WSNs) consist of distributed
autonomous sensor node that can be deployed to monitor the
inaccessible and accessible areas such as forest fire, deserts,
air pollution monitoring, deep oceans water quality
monitoring, glaciers. Each sensor network node generally
equipped with a radio transceiver with an internal or external
antenna ,memory Unit, micro controller, and each node
having own battery with limited energy to reducing the
power consumption of each node is a difficult task.
Fig.1. Heavy traffic-flow and roles of sensor nodes in WSN.
________________________________________________________________________________________________
ISSN (Print): 2278-5140, Volume-3, Issue-2, 2014
6
International Journal of Advanced Computer Engineering and Communication Technology (IJACECT)
________________________________________________________________________________________________
The all-node-active condition is not practical for energy
constraint WSN. The senor nodes are saving the energy by
active and sleep states. The total time of sensor nodes are
active and dormant state is called duty cycle. The duty cycle
depends on the network coverage and connectivity. Usually
for a dense WSN the duty cycle of a node is kept very low.
A duty cycle WSN can be loosely categorized into three
main types: random duty-cycled WSN, co-ordinated duty
cycled WSN, Adaptive duty-cycle WSN, in random duty
-cycle the sensor nodes can be turned active or sleep state
independently in random fashion. The random duty-cycle
required, but the disadvantage of random duty cycled WSN
will not go to the sleep state based on their network
condition.
It will be generating the heavy traffic. It will not use better
utilization of bandwidth. In coordinated duty cycle the
sensor node coordinates among themselves through the
communication and message exchange However, it requires
additional information exchange to broadcast the active
sleep schedules of each node. It will generate the heavy
traffic and overhead. This paper propose an adaptive duty
cycle control mechanism based on the queue management
with the aims of power saving and delay reduction. The
proposed scheme does not need explicit state information
from the neighboring nodes, but only uses the possessive
queue length available at the node. The network condition or
traffic variations changes implicitly occurs because the
queue states having possibility or power of the network
states. Using the queue length and its variations of a sensor
node, it present a design of distributed duty cycle controller.
Therefore the adaptive duty-cycled based WSN has been
considered for its design. Specifically the problem of
reduction of traffic in bottleneck zone has been considered.
Fig 2.Adaptive duty cycle in a WSN.
In this work, the primarily goal is to gain certain analytical
understanding on the upper-bound of the network lifetime
and comparative studies carried out to estimate upper bound
of the network lifetime between adaptive duty cycle WSN
without network coding and adaptive duty cycle WSN with
network coding technique.
A. Network Coding
resilience to attacks and eavesdropping. Instead of simply
relaying the packets of information they receive,
the nodes of a network take several packets and combine
them together for transmission. This can be used to attain the
maximum possible information flow in a network. The
intermediate nodes of a network can appropriately encode
the incoming data packets before forwarding the coded
packets to the next node reduces the traffic without
increasing delay. The network coding theory started with
R.Ahlswede, N. Cai, S. Y. R. Li, and R. Yeung seminal
paper [4]. The encoding and decoding methods of network
coding described below..
Fig 3.Encoding and Decoding operation of Sensor nodes.
Encoding operation: A node that wants to transmit encoded
packets, choose sequence of bits ‘b1’, ‘b2’ from nodes Ni
(i=1,2,3….n).The output encoded packet is given by
Y= Ni (b1⊕b2)
Where i=1,2,3…..n
(1)
The encoded packets are transmitted with n coefficient in the
network.
Decoding Operation: A node receives encoded packets, and
it will get bits either ‘b1’or ‘b2’ based on that it will
performing decoding operation A source node A send the
information to the destination F and G , the node A will be
generating information ‘b1’and ‘b2’ and it should be multi
casted . The node D receives the information of ‘b1’, ‘b2’ and
it will perform XOR operation. The Destination F receiving
the‘b1’bit and also get XOR result. It will again perform
XOR operation
b1⊕ (b1⊕b2 ) for getting b2 bit.
The XOR Network coding, a special case of linear network
coding [5], has been used. The fundamental idea of coding
in intermediate nodes in network has been shown to have
advantages in other scenarios such as minimizing network
resources, network diagnosis and packet communication in
wireless network. Network coding technique only needs a
few linear operations and several bytes storage to reduce
much power consumption, increasing the lifetime of sensor
networks.
Network coding is a technique which can be used to improve
a network's throughput, efficiency and scalability, as well as
________________________________________________________________________________________________
ISSN (Print): 2278-5140, Volume-3, Issue-2, 2014
7
International Journal of Advanced Computer Engineering and Communication Technology (IJACECT)
________________________________________________________________________________________________
II. LITERATURE SURVEY
Duty cycle facilitates in reduction of energy consumption in
a dense WSN. Furthermore, network coding technique has
been drawn its attention for improvement of throughput,
bandwidth and energy efficiency in resource constraint
wireless networks.
The Adaptive Duty cycled with queue management in WSN
has been estimated by Heejung Byun ,Junglok Yu [2]. Q.
Wang and T. Zhang [6] proposes that bottleneck zone in a
sensor network is considered as the intersection area
between the sensor deployment area and a bottleneck zone
centered at the sink. In a sensor network deployment, the
whole network relies on the nodes inside the bottleneck
zone to relay messages. Estimation of upper bounds of the
network lifetime through bottleneck zone analysis in (a)
random duty cycled WSN (b) non-duty cycled WSN using
network coding in the bottleneck zone (c) random
duty-cycled WSN using network coding in the bottleneck
zone has been done by R. R. R.R.Rout and S.K.Ghosh [1].
There have been studies on the network lifetime in WSNs.
Y.Wu,S.M.Das,and R.Chandra.[3] presented routing
protocols to aggressively exploit coding opportunities in the
network. The basic idea is to route the flows in the network
to a region where network coding can be performed. This
improves the throughput, however it greatly affects the
network lifetime. Concentrating large amounts of traffic to a
small region in the network burdens the involved nodes.
This routing approach can eventually lead to a network
breakdown. Therefore, such an approach cannot be applied
to energy constrained networks.
The network lifetime upper bounds in a cluster based WSN
has been estimated by S. Lee and S. H. Lee [7]. Bhardwaj et
al. [9] and Q. Wang and T. Zhang [6] have derived upper
bounds on network lifetime for a non-duty cycle based
WSN. A duty cycle based broadcasting scheme with
reliability has been proposed by F. Wang and J. Liu. [8]. A
random duty cycle based WSN has been considered for
dynamic coverage by C. F. Hsin and M. Liu. [10].A random
linear network coding based scheme that provides
packet-level capacity for both single unicast and single
multicast connections have been proposed by D. Lun, M.
Medard, R. Koetter, and M. Effros [11] for wireless
networks. R.R.Rout and S.K.Ghosh. [1] have also presented
a network coding based probabilistic routing scheme which
provides gains of network coding in a WSN.
The focus of the present work is to estimate the upper
bounds of network lifetime in WSN, considering (i) adaptive
duty cycle, (ii) network coding, and (iii) combinations of the
duty cycle and network coding. A network coding based
communication paradigm in the bottleneck zone has been
proposed to reduce the traffic load which enhances the
network lifetime. The theoretical analysis, simulation and
performance analysis have been done to show the efficacy of
the proposed approach.
III. ESTIMATION OF NETWORK LIFETIME
USING ADAPTIVE DUTY CYCLE
In this section system model has been explained. Based on
the system model, an energy consumption model for duty
cycle based WSN has been developed. The upper bound of
the network lifetime has been estimated and energy savings
due to duty cycle has also been shown.
A. System Model
A system is considered with N sensor nodes scattered
uniformly in area A. The area A with a bottleneck zone B
with radius D is shown in Fig. 1. Adaptive Duty Cycled in
WSN in Fig 2.All the N sensor nodes are duty cycle enabled
. The nodes are named based on their roles in the network as
shown in Fig. 1. The sensor nodes in the bottleneck zone
are divided into two groups: relay sensor and network coder
sensor nodes. The relay sensor nodes (R) transmit data,
which are generated outside as well as inside the bottleneck
zone. The network coder sensor nodes (N) encode the raw
native data which are coming from outside the zone B before
transmission. The sensor nodes outside the zone B are
marked as I and L in Fig. 1. The leaf sensor nodes (L)
periodically sense data and transmit them toward the Sink.
The intermediate sensor nodes (I) relay the data in the
direction of the Sink S. In the bottleneck zone, the relay
nodes can communicate with the Sink using a multihop
communication. However, the network coder nodes use a
single hop to communicate with the Sink. The radius, D,
should be at least equal to the maximum transmission range
of a sensor node, so that the data generated outside the
bottleneck zone can be relayed through the zone B.
Each sensor node has a number of Received queue and
sensed Queue attached to it, one or more to other nodes,
more to the sink. On each sensor node the packets are
arrived and depart except the Leaf (or) Terminal node and
Sink node. The proposed approach is to dedicate the buffer
at each node to a single FIFO queue. When the buffer
occupancy exceeds a threshold the switch begins to the
sensor node as an active state to do so until buffer occupancy
falls below the threshold. If the buffer size below the
threshold means the sensor node going to the dormant
(sleep) state.
B. Energy Consumption Model Using Adaptive duty cycle
A sensor node consumes energy at different states, such as,
sensing and generating data, transmitting, receiving and
sleeping state. Energy savings are done at the node level
through switching between active and sleep states. Energy
consumption by a source node per second across a distance d
with path loss exponent n is,
________________________________________________________________________________________________
ISSN (Print): 2278-5140, Volume-3, Issue-2, 2014
8
International Journal of Advanced Computer Engineering and Communication Technology (IJACECT)
________________________________________________________________________________________________
Etx = Rd (α11 + α2dn)
(2)
where ‘Rd’ is the transceiver relay data rate, ‘α11’ is the
energy consumed per bit by the transmitter electronics and
‘α2’ is the energy consumed per bit in the transmit op-amp.
Moreover, the total energy consumption in time ‘t’ by a
source node (leaf node) without acting as a relay
(intermediate node) is,
Es = t [p(rses + Etx) + (1 − p)Esleep]
(3)
where ‘Esleep’ is the sleep state energy consumption of a
sensor node per second, ‘rs’ is the average sensing rate of
each sensor node and it is same for all the nodes, ‘es’ is the
energy consumption of a node to sense a bit, the probability
‘p’ is the average proportion of time ‘t’ that the sensor node
devotes in active state. Thus, ‘p’ is the duty-cycle. A sensor
node remains in the sleep state with probability (1- p) till
time ‘t’. The energy consumption per second by an
intermediate node which act as a relay is given by
Etxr = Rd (α11 + α2 dn + α12)
= Rd (α1 + α2dn)
(4)
where‘α12’ is the energy consumed by the sensor node to
receive a bit. Total energy consumption till time ‘t’ by an
intermediate (relay) node is
ER = t [p(rses + Etxr) + (1 − p) Esleep]
(5)
TABLE I
Simulation Parameters
S.No
1
2
3
5
6
7
8
9
PARAMETERS
Number of nodes(N)
Area
Bottleneck Zone
radius(D)
Path loss
exponent(n)
α11
α12
α2
Esleep
Eb
10
Software
4
VALUES
103
200 x 200 m2
10 meters
2
0.937 μjoule per bit
0.787 μjoule per bit
0.0172 μjoule per bit
30 μjoule per second
25Kjoule
MATLAB R2013a
(8.1.0.604)
IV. ESTIMATION OF NETWORK LIFETIME
USING ADAPTIVE DUTY CYCLE AND
NETWORK CODING
region simply forward the data to sink. This relay node helps
the Sink to decode the encoded packets. Whenever a node in
the WSN, it will check the queue.
Adaptive Duty cycling and the packet processing in the
network coding layer of the bottleneck zone has been
explained below. Each node in the network maintains the
received queue (receiveQueue()) And a sensed queue
(senseQueue()). On sensing a information a node put the
packet in the senseQueue(Pj). On receiving a packet ‘Pi’ a
node put the packet in the receiveQueue(Pi). The node check
the queues length whether it is below the threshold or above.
If the threshold is below means the node gone to sleep state
otherwise It should be active state. If the received packet
already processed by the node than it is discarded, otherwise
node processes the packets further. The node check the role
from the Nodeset(), whether it is relay node or network
coder node. If the node is relay node in network coding layer
it will forward the packet to the sink otherwise forward the
packet to the network coding layer. The node is an encoder if
the packet received non coded packet then it performs the
XOR operation. On successfully generating encoded
packets, the network coding node transmits the encoded
packet to the Sink. The processed packets inserted into
Transmitted set Transmitset(). This stores the transmitted
packet and helps restricting further redundant transmission.
However the received packet ‘Pi’ is already processes then it
is discarded by node.
Decoding packet at the sink : The sink node receive the
packet from the relay nodes and encoded packets from the
network coding node. The decoding procedure is executed
at the sink which processes the gathered data in WSN. The
sink maintains a packet storage in which it stores the relayed
packets. And the sink receives an encoded packet containing
of ‘k’ relayed packet, the sink retrieves the relayed packet
one by one from the packet storage.The sink XORs the
received encoded packet with the (k-1) native packets to
retrieve the missing packet.
V. SIMULATION RESULTS
This section presents the simulation results of the proposed
algorithms. The energy efficiency and improvement of
lifetime of the proposed approach have been discussed.
A. Packet transmission via Relay Node
The Fig 4 shows that packet transmission via relay node. The
relay sensor nodes transmit data which are generated outside
as well as inside the bottleneck zone. The relay nodes simply
forward the received data.
In this section the network life time has been estimated with
a proposed network coding algorithm for a adaptive duty
cycled WSN. A network coding layer (refer Fig.1)
containing network coder nodes has been introduced around
the Sink. The network Coding layer is the most excessive
load in bottleneck region. So, the reduction of energy
consumption of the network coding layer leads to higher
network life time. The group of relay nodes in the bottleneck
________________________________________________________________________________________________
ISSN (Print): 2278-5140, Volume-3, Issue-2, 2014
9
International Journal of Advanced Computer Engineering and Communication Technology (IJACECT)
________________________________________________________________________________________________
network. It has been observed from Fig 6.that the network
lifetime with adaptive duty cycle and network coding is
more than using adaptive duty cycle. The improvement of
network lifetime is due to the introduction of network coding
nodes near the Sink. In Fig 7 energy consumption has been
shown for adaptive duty cycle without network coding and
with network coding. The per node energy consumption
with adaptive duty cycle is more than a with adaptive duty
cycle and network coding.
Fig.4 Packet transmission via relay node.
A. Packet transmission via Network Coder Node
The Fig 5 shows that packet transmission via network coder
node. Instead of simply relaying the packets of information
they receive, the nodes of a network take several packets
and combine them together for transmission. The incoming
data packets before forwarding the coded packets to next
node. This can be used to attain the maximum possible
information flow in a network.
Fig.6 Network Lifetime (a) with Adaptive Duty
Cycle(TuD) and (b) with Adaptive Duty cycle and network
coding (TuNCD).
With the use of MATLAB ,in this paper have studied
different parameters for two cases. Fig .6 and Fig.7 which
uses the parameter settings given in Table I and Table II
shows the comparison of simulation results.
TABLE II
S.
No
1
2
Shows the Comparison of Simulation Results
VALUES
VALUES
(for Adaptive
(for
PARAMETERS
duty cycle)
Adaptive
(If duty cycle
duty cycle
p= 5)
with
Network
Coding)
(If duty
cycle p= 5)
Network
Lifetime(Seconds)
0.2*e4
2.2*e4
Per node Energy
Consumption(Mic
3.0*e5
1.5*e5
ro Joules)
As the duty cycle increases from 1 to10 the number of
transmissions and receptions in the network increases. Thus,
the energy consumption of the nodes are also increases in the
Fig.7 Per node energy consumption (a) with Adaptive Duty
Cycle and (b)with Adaptive Duty cycle and network coding.
VI. CONCULSION
In a wireless sensor network the area near the sink forms the
bottleneck zone where the traffic flow maximum, so
reducing the power consumption. In this paper, propose
adaptive duty cycle control for WSN and network coding.
The proposed approach controls the Adaptive duty cycle
through the queue management to achieve high performance
under network condition changes. The sensor node save the
energy by dynamically changing the sleep time based on the
predetermined value of queue length.
These results in minimum power consumption and faster
adaptation to traffic changes. The proposed scheme only
________________________________________________________________________________________________
ISSN (Print): 2278-5140, Volume-3, Issue-2, 2014
10
International Journal of Advanced Computer Engineering and Communication Technology (IJACECT)
________________________________________________________________________________________________
requires the local queue length for computing the duty cycle,
which adds good scalability, improved efficiency to the
system. In addition, propose a network coding approach
which is used to improve the capacity of information flow
with the better utilization of bandwidth in a multi hop
communication.
[4]
R. Ahlswede, N. Cai, S. Y. R. Li, and R. Yeung,
“Network information flow,” IEEE Trans. Inf.
Theory, vol. 46, no. 4, pp. 1204–1216, July 2000.
[5]
[9] S.-Y. Li, R. W. Yeung, and N. Cai, “Linear
network coding,” IEEE Trans. Inf. Theory, vol. 49,
no. 2, pp. 371–381, 2003.
[6]
Q. Wang and T. Zhang, “Bottleneck zone analysis in
energy constrained wireless sensor networks,” IEEE
Communications Letter, Vol. 13, No. 6, pp. 423–
425, June 2009.
[7]
S. Lee and S. H. Lee, “Analysis of network lifetime in
cluster-based sensor networks,” IEEE Commun.
Lett., vol. 14, no. 10, pp. 900–902, 2010.
[8]
F. Wang and J. Liu, “RBS: a reliable broadcast
service for large-scale low duty-cycled wireless
sensor networks,” in Proc. 2008 IEEE ICC, pp.
2416–2420.
[9]
Heejung Byun, Junglok Yu, "Adaptive Duty Cycle
Control with Queue Management in Wireless Sensor
Networks,"IEEE
Transactions
on
Mobile
Computing, vol. 12, no. 6, pp. 1214-1224, June
2013, doi:10.1109/TMC.2012.102
M. Bhardwaj, T. Garnett, and A. Chandrakasan,
“Upper bounds on the lifetime of sensor networks,”
in Proc. 2001 IEEE ICC, pp. 785–790.
[10]
C. F. Hsin and M. Liu, “Randomly duty-cycled
wireless sensor networks: dynamic of coverage,”
IEEE Trans. Wireless Commun., vol. 5, no. 11, pp.
3182–3192, 2006.
Y. Wu, S.M. Das, and R. Chandra, "Routing with a
Markovian Metric to Promote Local Mixing,"' MSR
Technical Report, Nov 2006.
[11]
D. Lun, M. Medard, R. Koetter, and M. Effros, “On
coding for reliable communication over packet
networks,” Physical Commun., vol. 1, pp. 3–20,
2008
Network coding is not simply relaying the packets of
information they receive, the sensor nodes of a network take
several packets and combine them together for transmission
and applied in bottleneck zone. That the proposed algorithm
improves significantly both energy efficiency and delay
performance by adapting the duty cycle properly under
network changes and derived performance figures using
MATLAB.
REFERENCES
[1]
[2]
[3]
R.R.Rout and S.K.Ghosh,"Enhancement of lifetime
using duty cycle and Network Coding in wireless
Sensor Network,"IEEE Transactions on Wireless
Communications, Vol.12, No.2, pp.656-667,
February 2013.

________________________________________________________________________________________________
ISSN (Print): 2278-5140, Volume-3, Issue-2, 2014
11