Wireless Sensor Network, 2013, 5, 41-51
http://dx.doi.org/10.4236/wsn.2013.53006 Published Online March 2013 (http://www.scirp.org/journal/wsn)
A Survey on Modeling and Enhancing Reliability of
Wireless Sensor Network
Latha Venkatesan1, S. Shanmugavel2, Chandrasekaran Subramaniam1
1Velammal Engineering College, Anna University, Chennai, India
2Anna University, Chennai, India
Received December 17, 2012; revised January 18, 2013; accepted January 28, 2013
ABSTRACT
Design of reliable wireless sensor network (WSN) need to address the failure of single or multiple network components
and implementation of the techniques to tolerate the faults occurred at various levels. The issues and requirements of
reliability improvement mechanism depend on the available resources and application for which the WSN is deployed.
This paper discusses the different modeling approaches to evaluate the reliability and classification of the approaches to
improve it. Also the paper analyzes reliability enhancement by existing fault tolerant methods in WSN and compares
the performance of these techniques with the technique we developed. From the results of the analysis we highlight the
challenges and the characteristics of the sensor network affects the reliability and give some scope of future research
directions in order to enhance reliability.
Keywords: Component Reliability Modeling; Reliability Measure; Reliability Requirements; Fault Tolerance; Quality
of Information
1. Introduction
A wireless sensor network is a collection of large number
of sensor nodes deployed over the region or inside the
target which is to be detected, monitored or tracked. Each
sensor node consists of processing capability, memory, a
RF transceiver, a power source, and accommodate vari-
ous sensors and actuators [1]. These nodes self organize
into a cooperative network [2] to communicate in an ad
hoc fashion and transmit the sensor measurements to the
end user. Such systems can revolutionize the way we live
and work. Currently, wireless sensor networks are begin-
ning to be deployed at an accelerated pace. It is not un-
reasonable to expect that in 10 - 15 years that the world
will be covered with wireless sensor networks with ac-
cess to them via the Internet. Many applications have
been proposed including environmental, medical, mili-
tary, transportation, entertainment, crisis management,
homeland defense, and smart spaces [3]. These applica-
tions may require reliable network for collecting all data
without loss from nodes. But on the other hand the inex-
pensive sensor nodes may not be highly reliable since
they are limited in energy, memory space and proces-
sing capabilities, and the onboard sensors have direct
contact with the environment. This results in error intro-
duced in some of the sensor measurements while sensing,
processing or reporting the data to the sink. So in order
to improve data integrity and detection reliability [4],
the faulty sensor data have to be detected and filtered out
while data without fault are aggregated and sent. Even
when the object or event is reliably detected, error may
be introduced in mutihop communication due to poor
link quality and to improve the network based reliability
can be by suitable routing protocol. Also even if the de-
vices and links are working properly, sometimes the
event or targets may not get detected due to limited cov-
erage and connectivity and inefficient placement of sen-
sor nodes.
The first task of reliability improvement is to specify
reliability requirements as defined by the network main-
tenance operator (QoS) and/ or end users in terms of ope-
rational performance reliability requirements and the
operating environment. Reliability requirements address
specifications, testing, assessment of the system and ser-
vices provided by the system. Without defining the re-
quirements the network will not serve its application re-
liably and also the resources will be wasted. Some com-
mon examples of reliability requirements metrics are [5]
MTTF (Mean Life or Mean-Time-To-Failure), MTBF
(Mean-Time-Between-Failures), failure rate. Once the
requirements have been established determine reliability
improvement approaches and apply them and measuring
the improvement of the performance is the second task.
The next task in the reliability improvement process is
reliability modeling and testing, to assess its reliability
C
opyright © 2013 SciRes. WSN

L. VENKATESAN ET AL. 42
level. In the early phase for example during the deploy-
ment, reliability modeling can be used to predict the
network performance to provide information for suitable
position of the sensor nodes. Reliability modeling and
analysis are key steps to the design and optimization of
sensor network systems. There has been little work done
in the reliability modeling and analysis of WSN.
The reliability improvement process is an iterative
process that is applied step by step in each stage of im-
plementation of WSN. It can be improved one or multi-
ple components like hardware, software or any ser-
vice/functions of the network in order to achieve mini-
mum required reliability. One of the approaches to im-
prove the reliability of a system is fault tolerance (FT)
defined as tolerating the faults which may occur during
node placement, topology control, target or event detec-
tion, data aggregation, routing and information process-
ing. Fault tolerance, is achieved by hardware or informa-
tion redundancy. Usually redundancy can result in in-
creased design complexity and increased costs but re-
dundancy is an inherent property of the sensor network.
Nodes in WSNs are prone to be failure due to energy
depletion, hardware failure, communication link errors,
malicious attack etc [6] can be tolerated by hardware and
software redundancy, which in turn improves the inher-
ent reliability, and information redundancy and time re-
dundancy will improve the operational redundancy. In
general statistical reliability with level β [8] of WSN is a
QoS (Quality of Service) level where during every pre-
determined time window, a predetermined size of ran-
dom sensed data are delivered to the sink node from
every source, each with a probability of at least β. But
studying the characteristics, and considering constraints
of WSN, it is understood that the overall reliability de-
pends on many aspects spans from reliable sensing to
receiving information in the sink and application under-
lying.
Most of research works with the objective of reliability
improvement addresses network transport protocol and
improve end-to-end reliability. Reliability improvement
is achieved by improving any one or two aspects with
minimal consideration of characteristics of the WSN. In
this paper we attempt to examine the existing reliability
enhancement techniques which will tolerate some faults
and compare them based on their performance. The re-
mainder of the paper is organized as follows: In Section
2 the aspects previous survey works are discussed, in
Section 3 methods used for modeling and analyzing the
reliability of WSN and its classification are explained. In
Section 4 the factors influence the reliability of WSN are
investigated, in Section 5 various existing reliability en-
hancement approaches are compared based on the per-
formance and conclusions are made in Section 6.
2. Previous Work
In WSN a large number of sensor nodes continually
sense data from the environment and the critical event
data need to be reliably delivered to the sink. The sink
receives all the information from these sensor nodes,
processes it and sends them to the end user. Therefore,
given the nature of error prone wireless links, presence of
moving nodes and failing nodes, ensuring reliable trans-
fer of data from resource constrained sensor nodes to the
sink is one of the major challenges in WSNs. There is
extensive research done on reliable transport protocol
and survey on existing data transport reliability protocols
in wireless sensor networks is presented in [9,13,14]. The
network may be congested and sensor nodes may drop
packet when the data packets generated by a large num-
ber of sensor nodes exceed the network capacity. In order
to achieve reliability, the dropped packets must be re-
transmitted which in turn leads to wastage of energy and
bandwidth, very important stringent factors in WSN.
Thus the reliability enhancement is achieved with the
trade between reliable data transmission and network
resources. In order to mitigate this problem, the reliable
transport protocol consists of two essential mechanisms:
congestion control (detection and avoidance), and loss
detection and recovery with the minimum possible en-
ergy consumption. A Comparison between the protocols
based on the method by which it recover the losses (End
to End recovery, Hop to Hop recovery), messages to de-
tect losses control congestion, type and level of reliability,
energy efficiency, packet-driven reliability or event-
driven reliability. Also a classification of reliability as
Packet or event reliability is concerned with how much
of information is required to notify the sink of an occur-
rence of something happening in the environment.
It is well accepted that a sensor network should be de-
ployed with high density, in order to prolong the network
lifetime. Sensor data collected from such network is
likely to be highly correlated and redundant, so a func-
tion density control controls the density of the active
sensors to certain level. At any instant of time only a
subset of sensor nodes will be active, sensed and report
to sink while other sensors are inactive must guarantee
sufficient sensing coverage and connectivity reliabilities.
The performance metrics of interest are 1) the percentage
of coverage, i.e. the ratio of the covered area to the total
area to be monitored; 2) the number of working nodes
required to provide the percentage of coverage in 1); and
3) α-lifetime, defined as the total time during which at
least α portion of the total area is covered by at least one
node. The ability to report the Sink node is called as
connectivity. A network is said to be fully connected if
every pair of node can be communicated with each other
either directly or via intermediately relay nodes. It is im-
portant to find the minimum number of sensors for a

L. VENKATESAN ET AL. 43
WSN to achieve the connectivity. The connectivity of a
graph is minimum number of nodes that must be re-
moved in order to portion the graph in to more than one
connected component. Connectivity affects the robust-
ness and throughput of the wireless sensor network. Both
the coverage and connectivity are related to each other
for improving the performance of Wireless sensor net-
works. The issues in maintaining sensing coverage and
connectivity by keeping a minimum number of sensor
nodes in the active mode in wireless sensor networks are
presented in [15,17]. Various approaches to model the
coverage are compared based on how they address the
issues in coverage; some of them are coverage types,
deployment strategy, sensing model, sensing area and
nature of the algorithm. The system parameters, such as
the initial energy of a node, the radio transmission rate,
and the energy consumption rate, are assumed to same
for all the nodes.
Sensor network is fault prone in the sense that 1) Con-
sists of low cost, less reliable sensor nodes with limited
energy, 2) Sensor nodes are deployed in the unattended
hostile environment, 3) Erroneous communication
through Low quality Wireless links, 4) Multihop com-
munication between node and sink, leads to low per-
formance and reliability. The nature of faults occurred in
the WSN is different from other networks. Additionally,
manual inspection of faulty sensor nodes after deploy-
ment is typically impractical. Nevertheless, many WSN
applications are mission-critical, requiring continuous
operation. Thus, in order to meet application require-
ments reliably, WSN must have the ability to detect and
recover from faults, the two mechanisms of FT. The ob-
jective of fault detection is to provide any countermea-
sures, the first step a system must perform is to detect
that a specific functionality is or will be faulty. Fault
Recovery is, after the system has detected a fault, the
next step is to prevent or recover from it. Enhancing the
service availability and reliability in WSNs through the
use of fault tolerance techniques is presented in [20,21].
These papers explains the taxonomy of classification of
faults and the corresponding failure are defined and the
methodology to detect the faults and how redundancy is
used to recover from failure. The other two techniques of
FT are identification and isolation of faults are explained
in [22] with respect to WSN. The paper [23] present and
classify various approaches for redundancy in the area of
wireless sensor networks, related to sensing, communica-
tion and information processing.
3. Reliability Modelling and Analysis
Methods
Reliability is one of the most important performance
measures and high level of reliability is a significant re-
quirement for a wireless sensor networks using in Indus-
trial and medical environments. Reliability level of the
network can evaluated using the tool called as reliability
modeling. Reliability modeling and analysis are key
steps to the design and optimization of sensor network
systems. The approach generally taken to investigate the
reliability of a highly reliable WSN is
1) Develop a mathematical model of the reliability
measure of the network;
2) Measure or estimate the parameters of the model;
3) Compute the network reliability based upon the
model and the specified parameters.
During the deployment of WSN, reliability modeling
can be used to predict the performance of the network
elements to provide the information for the design of
WSN suitable for an application in hand. For a network
already deployed in the field reliability modeling com-
bined with failure data analysis, can be used to identify
the critical components, apply Fault tolerance and en-
hance reliability.
3.1. Reliability Measure
The There exist a number of reliability measures for a
wireless network depending on the network and its ap-
plications. For a telecommunication network it is obvi-
ously communication issues that meet certain connec-
tivity requirements. Whereas for the sensor network the
focus is on information gathering, processing and com-
munication issues to meet the coverage and connectivity
requirements. Different criteria can be considered in or-
der to express or measure the reliability of a network.
The main ones are the following:
 Reliability measure of connectivity falls within any
one of the categories, which are 2-terminal reliability,
k-terminal reliability, all terminal reliability and many
sources to terminal reliability. For example 2-terminal
reliability can be computed using computationally ef-
ficient source to terminal path sets or cut sets enu-
meration algorithm;
 Hardware reliability measure are MTTR and MTTF;
 The coverage reliability measure is guaranteed de-
sired level of coverage (at least k-coverage) of event
or target at all times formulated by either Boolean
sensing model or collaborative sensing model [25];
 Capacity/Max flow measure is defined as the prob-
ability that the maximum flow of the network is not
less than the given demand.
 QoS reliability measure is guaranteed date transfer
timely and guaranteed bandwidth data accuracy time-
ly depends upon user/applications demand [26];
 Information reliability ensures that the nodes transmit
to the sink only information concerning significant
events or targets.
Each reliability measure is concerned with the ability
of a network to be available to provide the desired ser-

L. VENKATESAN ET AL. 44
vice to the end user.
3.2. Classification of Reliability Modeling and
Evaluation
Reliability of WSN, depends on combination of hard-
ware, software and wireless link is modeled in many
ways analogous to reliability modeling of hardware sys-
tem. Reliability modeling aims at using abstract repre-
sentation of the network as means for assessing their re-
liability. The basic reliability modeling techniques avail-
able in the existing work in WSN field are classified as
shown in Figure 1.
Two of the most commonly used Deterministic reli-
ability modeling and analysis methods are First Order
Reliability Method (FORM) and the Second Order Reli-
ability Method (SORM). The advantages of the tech-
niques is less computational effort than Probabilistic
models, but not suitable for modeling the reliability of
real time applications of WSN. In a more restrictive
sense, the term reliability is defined to be the probability
that a wireless sensor network performs its mission suc-
cessfully. Because the mission is often specified in terms
of time, reliability is often defined as the probability that
a system will operate satisfactorily for a given period of
time. Thus reliability may be a function of time and can
be estimated essentially using probability modeling.
When the set of operating components and the set of
failed components of a network are specified, it is possi-
ble to compute the probability that the system is operat-
ing and thus the reliability of the system. In probabilistic
modeling, the concepts and methods of probability the-
ory to compute the reliability of a complex system like
sensor network.
Combinatorial model: It can be used to analyze tran-
sient or steady state network. These model types are
similar in that they capture conditions that make a system
Figure 1. Classification of reliability models.
fail in terms of the structural relationships between the
system components. Various categories exists under
combinatorial Model, some of them that are commonly
used for WSN are reliability block diagrams, reliability
graphs and fault trees. Reliability block diagrams (RBD)
model has been widely used as one of the most practical
reliability modeling tools due to its simplicity. RBD
model consists of an input point, an output point, and set
of components are combined into blocks in series, in
parallel or in k-out-of-n configurations. Each block
represents a physical component that functions correctly.
The blocks in the RBD are arranged in a proper combi-
nation of series and parallel or k-out-of-n such a way that
illustrates working components that keep the entire sys-
tem operational. RBD for data transport in the WSN is
discussed in [33] the reliability of data transport is mo-
deled. The Fault tree is a pictorial representation of the
combination of events that can cause the occurrence of
an undesirable event. It represents all the sequences of
individual component failures that cause the system to
stop functioning, in a treelike structure. Traditional fault
trees can only express the system failure in terms of the
combination of component failures through the use of
AND, OR, and n-out-of-m logic gates. The occurrence of
each event is denoted by a logical 1 at that node; other-
wise the logic value of a node is 0. The process of build-
ing a fault tree is performed deductively and starts by
defining the TOP event, which represents the system
failure condition. From this event, and by proceeding
backwards, the possible root causes are identified. The
events at the bottom of the tree are referred as basic
events. If a basic event occurs two or more times in a
fault tree it is called a repeated event. Some of the reduc-
tion algorithms of Series-parallel formula, VT algorithm
the factoring/conditioning algorithms allow to resolve
reliability problem in linear time for repeated event fault
tree. A three-level hierarchical model for the reliability
analysis of a sensor network is given in [34] as top model
for the sensor network, the middle level model for sensor
node are the fault tree and the bottom level models are
Markov chains for individual sensor node components
such as the power, sensor, ADC, microcontroller, exter-
nal memory), tiny OS and channel model. The work in
[35] focus on the approach called as sum of disjoint
products (SDP) efficiently in fault trees consists of re-
peated events and it is easily automated in sensor net-
work. The expressive power of the fault tree can be ex-
panded by defining some dynamic gates for describing
the modular imperfect coverage behaviour and com-
mon-cause failure behaviour of WSN [36]. Basically, a
fault tree can be used to model a system with only per-
manent faults (no transient or intermittent faults) and
reconfiguration is not possible. Also assume component
failure is independent and state-dependent behaviour

L. VENKATESAN ET AL. 45
cannot be modelled. Because fault trees are easier to
solve than Markov models, fault trees should be used
wherever these fundamental assumptions are not violated.
Fault tree and reliability block diagram are called as non
state space models.
Path Enumeration methods: A reliability graph is
equivalent to a non-series-parallel reliability block dia-
gram and directly analyzed by many algorithms to re-
solve the reliability problem. Any communication net-
work can be considered as directed graph G = (V, E),
where V—vertices represents the sensor nodes and sinks,
Eedges represents the link between the two nodes.
Graphical modeling is the exact method in which there
are two classes for the computation of the network reli-
ability, which are path or cut enumeration methods and
network reduction methods. In path enumeration method,
first step is the enumeration of all the minimum paths for
a working network to provide a Boolean expression
called as Structure function. These minimal paths are
different for each reliability problem. For example two
terminal reliability of the network is the probability that
there exist atleast one or more successful paths from the
target sensor node to the sink. The second step is com-
putation of this Boolean expression probability. The
computational problem to determine the complete set of
minimal paths or cuts and to construct and evaluate the
structure and reliability functions is difficult. In general,
if there are k cuts or paths, the corresponding reliability
equation will have (2k − 1) terms. Several techniques
available to reduce the structure function and compute
the reliability in polynomial time. WSN application for
structure health monitoring can represented by a proposi-
tional directed acyclic graph (PDAG) through which the
reliability is evaluated in three steps [49]. In first step the
connectivity information of the network is represented in
the matrix form and using LDF algorithm, the structure
function is determined. The structure function consists of
connectivity to node and link connected using logical
AND, OR operators. In the second step the complex
network structure function is reduced by introducing the
context sensitive information of sensor network. The
context of a wireless sensor network may be expressed in
terms of the network elements or the conditions in which
the sensor nodes are deployed. The third step is computa-
tion of network reliability for the reduced context aware
structure function is obtained by assigning the probabili-
ties of connectivity to the nodes and the edges. The re-
duced expression consists of minpaths from source to
sink, connected by logical operators and the reliability of
the WS network is the reliability of that min path which
has the maximum reliability out of many min paths.
In the network reduction algorithms are based on
graph topology, reduce the size of the graph by removing
some structures. Then it is possible to compute the reli-
ability in linear time and the reduction will result in a
single edge. The Binary Decision Diagram (BDD) struc-
ture provides compact representation of Boolean expres-
sions and is used to work out the terminal reliability of
the links. Set of algorithms for efficient construction and
manipulation of BDD structure. An enhanced ordered
binary decision diagram (OBDD) algorithm is proposed
in [37] to evaluate the reliability of wireless sensor net-
works (WSNs), based on the considerations of the com-
mon cause failure (CCF) and a large number of nodes in
WSNs. It is also shown that OBDD can decrease the cost
of OBDD constructions and storage. Decomposition al-
gorithms compute k-terminal reliability in linear time by
decomposing the graph into sub graphs. k-coverage and
k-connectivity reliability can be computed as the prod-
uct of coverage or connectivity probability of all nodes
and links in the path [27,28]. The performance metrics of
WSN-Packet delivery ratio [29], message delay [30] with
a reasonable overhead (in terms of retransmissions, ac-
knowledgement messages, and control messages) is also
measured. For computation of reliability the probability
of link existence and failure rate of node are not only
considered but also reliability of a node depends on
various environmental parameters such as noise, tem-
perature, pressure and magnetic effects are also consid-
ered. The development of reliability model needs to take
in account all of these parameters [31].
State enumeration methods: A very basic method of
Markov modeling to compute the sensor network reli-
ability involve enumerating all possible states of the
network based on the states of its components, all possi-
ble transition between those state and rate parameters of
those transitions. In reliability modeling the possible
states are healthy and failed. The reliability of fault tol-
erant WSN is modeled and analyzed in [38-40] using
Markov modeling technique A network with “n” number
of elements there are 2n possible states termed as state
space explosion. Thus it becomes computational unfeasi-
ble for a network with large number of elements since
the network reliability problem is NP-hard. So the states
that results in successful network operations are identi-
fied in Hidden Markov model and the probability of oc-
currence of each successful state is calculated. The reli-
ability of the network is the sum of all successful state
probabilities. Markov reward modeling is more powerful
techniques suitable for reconfigurable and fault tolerant
networks with redundancies and failure dependencies. A
Generalized Stochastic Petri nets (GSPN) model is de-
signed to evaluate the reliability of a parallel-redundant
fault tolerant WSN. Unlike Markov model, GSPN is
more observable, scalable and portable. The enumeration
methods provides an exact solution of reliability prob-
lems, yet the method becomes computationally expensive
for the analysis of large networks since the number of

L. VENKATESAN ET AL. 46
possible configurations increases exponentially with the
number of mobile nodes. So the enumeration methods
can be employed for networks of relatively small size
and there is a need of faster computational procedures
that can provide an accurate approximation of complex
WSN. In [43] paper, the author present a hexagon tessel-
lation sensor network model with role assignment
scheme and estimate the reliability and lifetime distribu-
tion by employing a new improved Monte Carlo scheme
which incorporates both simulation and analytic methods
and is suitable for the WSN. Also it is shown that the
node sub-roles and node density have important effects
on the network reliability and lifetime. A new concept
known as event reliability is defined in [44] and the au-
thor propose Event Reliability Protocol (ERP), an
event-to-sink reliability protocol that serves to improve
the scalability of event detection in a WSN by minimiz-
ing the unnecessary retransmission of data packets com-
ing from multiple nodes in an event’s locality. The per-
formance of the protocol is simulated and Event reliabi-
lity is analysed using a network simulator GloMoSim.
Characteristic aspects of sensor networks such as li-
mited accessibility and dynamic topology changes com-
plicate the usage of traditional reliability evaluation
methods. The limitations of the traditional models over-
come by presenting [45] a model-based approach to as-
sess the reliability of redundant sensor networks.
Weibull failure model can be used in reliability theory
to model the time to failure or life time modelling pre-
sented in [46] to examine the coverage and reliability of
distributed sensor networks that are designed for surveil-
lance applications. In Weibull distribution the failure rate
is proportional to a power of time and it can be used to
model life time of WSN defined as the time to first fai-
lure. In the reliability modeling field, we sometimes en-
counter systems with uncertain structures, and the use of
fault trees and reliability diagrams is not possible. To
overcome this problem, Bayesian approaches offer a
considerable efficiency in this context. The paper [47]
introduces recent contributions in the field of reliability
modeling with the Bayesian network approach. Bayesian
reliability models are applied to systems with Weibull
distribution of failure. The advantages of this modeling
approach are presented in the case of systems with an
unknown reliability structure, those with a common
cause of failures and redundant ones.
4. Factors Affecting Reliability in WSN
A basic way to improve the reliability is to have redun-
dant components; it may be hardware, software or infor-
mation. Reliability analysis of the WSN provides a
measure of the performance of the WSN. As the impor-
tance of reliability in WSN is discussed in the previous
section, this section it is appropriate to investigate the
factors affecting reliability so that the reliability can be
improved by improving these parameters. In order to
ensure that the network supports the application’s re-
quirements, it is important to understand how each of the
categories of wireless sensor networking affects reliabili-
ty. Challenges to achieving reliability on Wireless Sensor
Networks can be divided to four main categories which
are network elements, networking characteristics, condi-
tions in which WSN is deployed and strategies used to
design the network for an application are represented in
Figure 2.
The first kind of problems comes from the limited re-
sources of Wireless Sensor Networks nodes. A sensor
node has a limited power without recharging capability
affects the hardware reliability. The node is able to sense
the environment using sensors and ADC used to convert
to digital signal has low accuracy and low resolution will
affect the information reliability. Similarly the processor
with low computational power and memory space affects
the Quality of Information (QoI).
Next kind of problems comes from software engineer-
ing point of view adds more challenges to achieve reli-
ability of sensor network. In general, sensor networks
software can be layered into levels: sensor node software,
middleware and sensor network software. The sensor
node software is the bottom layer contains sensor soft-
ware and node software. The sensor software (Operating
System) has full access to the sensor hardware, execute
the process by which events occurred in the real world is
sampled and converted into machine-readable signals.
Malfunction or bucks in the software or programs will
affect the QoS and information reliability. The node
software process the raw data, receive and process the
query and transmit the data to toe next node includes
system software for network maintenance and applica-
tion specific software related with packet reliability.
Figure 2. Factors Influence the reliability in WSN.

L. VENKATESAN ET AL. 47
Nextica-
re related to the networking
ch
by the environmental
co
s related to strategies used to build
th
less communications, re-
so
5. Reliability Improvement in WSN
twork reli-
level, a collection of common services for appl
tion development, called middleware; reside over the
operating system related with service reliability. Appli-
cation programs use this middleware according to their
own specific requirements; these programs often access
the individual node resources and local services. Finally,
the sensor network software specifies the main tasks and
required services of transporting the data and query
through the network, maintain the performance and
achieve network reliability.
Third sort of problems a
aracteristics and nature of wireless communication link.
The asymmetry of links makes link quality estimation
hard and invalidates many assumptions made in other
environments. Correlated losses due to obstacles, mobi-
lity and interference can lead to consecutive losses, de-
creasing the effectiveness of erasure code. Weak correla-
tion between quality and distance, hidden terminal prob-
lems, and dynamic change of connectivity complicates
the situation further [48]. In addition, as its communica-
tion bandwidth is narrow, overhead due to the error cor-
recting coding cannot be added with the data packet. All
these features of wireless communication link degrade
the transport or packet reliability.
Fourth kind of problems raised
nditions affects the sensing unit and wireless trans-
ceiver components of a sensor node since they directly
interact with the environment, which is subject to variety
of physical, chemical, and biological factors. The prob-
ability that a sensor suffers a physical attack in such an
environment is therefore much higher. It results in low
reliability of performance of sensor nodes. Even if condi-
tion of the hardware is good, the communication between
sensor nodes is affected by many factors, such as signal
Strength, antenna angle, obstacles, weather conditions,
and interference [50].
Fifth set of problem
e application which includes deployment, scalability,
security and responsiveness. Deployment strategy is an
important issue to place the sensor nodes, relay nodes,
base station and sink in the proper position to maximize
coverage, connectivity, detection and communication
reliability and thus improve the life time of the sensor
network. Given the area to be covered, energy, range of
sensing and transmission radius of the node, many tech-
niques are available for calculating the number and posi-
tion of the nodes. Many aspects of WSN like topology
management, power management, and routing requires
suitable placing of the nodes deployment problem can be
stated in many ways and there is no one systematic
methodology for solving the deployment problem. But
improper deployment may hamper the service provided
by the sensor network which leads decrement the overall
network reliability. In typical wireless ad hoc networks,
reliability and scalability are always inversely coupled.
In other words, it becomes more difficult to build a reli-
able ad hoc network as the number of nodes increases.
This is due to the network overhead that comes with the
increased size of the network.
Because of the nature of wire
urce limitation on sensor nodes, size and density of the
networks, unknown topology prior to deployment, and
high risk of physical attacks to unattended sensors, it is a
challenge to provide security in WSNs. Security attacks
modify the information by packet damage and render the
information unavailable by dropping the packets affects
the information and packet reliability. Wireless networks
are vulnerable to security attacks due to nature of wire-
less communications, resource limitation on sensor nodes,
size and density of the networks, dynamic topology and
the attackers may device different types of security
threats to make the WSN system unstable. Typical treats
and adequate defense techniques in WSNs are summa-
rized in [51]. The security mechanisms are actually used
to detect, prevent and recover from the security attacks.
A wide variety of security schemes to counter malicious
attacks and these can be categorized and listed in [52].
But the question is how much the security mechanisms
are secure? The reliability of security systems is con-
nected with proper implementation of software and
hardware components of security systems. When security
of information systems is considered it is needed to ana-
lyze three attributes: confidentiality, availability and in-
tegrity. If these goals are not met then no proper routing,
clustering, aggregation process carried over in a WSN
and the reliability will certainly decreases. Responsive-
ness is the ability of the network to quickly adapt itself to
changes in topology. To achieve high responsiveness, an
ad hoc network should issue and exchange more control
packets, which will naturally result in less scalability and
less reliability.
The approaches proposed to reinforce the ne
ability, provide a method for maximizing reliability un-
der certain given constraints and WSN has sensor node
constraints like battery power, transmission range, sens-
ing range and processor capability. In addition network
constraints are ad hoc nature, data rate, packet size,
wireless link, scalability, security, dynamic topology and
cost constraints. For WSN the focus is on sensing, proc-
essing and communication issues that meet some reli-
ability measures and requirements. The communication
issue has to meet the connectivity requirements so that
the end user is able to access the network and receive the
information sensed and processed by the sensor nodes.
Achieving the overall network reliability of the commu-
nication process is to construct a network the minimum

L. VENKATESAN ET AL. 48
tworking constraint and can be meet
Table 1. Comparison of reliability improvement techniques.
Name of the protocol/ Reliability enhanced Constraints satisfied Performance improvedTechniques used
number of reliable links and each link must be economi-
cally feasible. The transport layer of WSN provides
mechanisms for establishing a connection between sen-
sor network and end users, while supporting Quality of
Service (QoS) mechanisms to ensure the properties of the
link lives up to a required quality of connection and reli-
ability. In general, there are several approaches to
achieve reliability against unreliable link, e.g., automatic
repeat request, multi-path routing and source coding. On
the other side the overhead increases and network be-
come overloaded. That is, these traditional approaches
are not suitable for WSN, many research works concen-
trate in improving the existing transport layer protocols
or developing new efficient reliable data transfer scheme
that satisfy the constraints and improve the performance.
To cope with the above issues, many works has been
proposed for efficient reliable data transfer scheme, some
of them are network coding [53], cross-layer strategy that
considers physical layer (i.e., power control), MAC layer
(i.e., retransmission control) and network layer (i.e.,
routing protocol) jointly [54], hybrid method based on
multipath data sending [55] and behavior-based trust
mechanism [56].
Scalability is ne
out in WSN through efficient clustering. The cluster head
node is used to receive data from other nodes in the clus-
ter, aggregate it and then transmitting the processed in-
formation. Data aggregation is a process whereby data
from sensors in an area is gathered for performing fusion
to eliminate redundant data to provide the useful fused
information to the base station and improve reliability
and life time of the network. The clustering and cluster
head selection process involves many issues of power
efficiency [57], traffic distribution, link quality distance
to sink etc to guarantee reliability. Aggregated values
when transmitted to sink, aggregated values from other
clusters can be fused using in the intermediate nodes to
reduce traffic in WSN. Rule based Fuzzy logic, trust
based, weight based, rough set theory are some of the
secure and reliable techniques to tolerate faulty nodes
and improve the credibility of the fused content in the
presence malicious nodes.
Sensors deployed in the real world senses the physical
parameters include temperature, acoustics, light and pol-
lution. Sensor network in the war field need to detect,
classify and track the objects in the vicinity. The reliabil-
ity can be defines as the successful detection by a node to
correctly estimate a target’s presence while avoiding
technique
E2SRT [32] Event-to-sink
reliability
Dynamic topology,
unreliable link Energy efficiency Includiol and ng congestion contr reducing convergence time
GARUDA [10] Multiple reliability
retransead Latency, energy efficiencyLo
RCRT [11] End-to-end reliability h
Implementing NACK based end-to-end
ZARB [12] Packet reliability M Network efActiets
HERO [19] Bi-directional pacNode and network Energy efe Efficient cluster
Data fusion with [41] Tranand
in
Unresion Energy efFusion depen
Improved CICADA [28] End-to-end Mobility
T
Randomization and overhearing
Simple CRT based
p] Packet forwarding
Unreliable
chaic
topoC
Power saving, simplicity,Packet-splitting algorithm
baer
EIRDA [24] Clustering Security Energy efficiency, life time,
Functional reputations for sensing,
a
CAP [7] Event reporting Node constraint Fault tolerance
Collaborative
distributed detection [16] Information reliability Faulty nodes Energy efficiency, life time, fault tolerance
Polynomial, aggregation the median
information, and judges the final
state of the event-region
semantics
Packet size,
mission overh
ss recovery through minimum
dominating set
Asymmetric link,
ardware constraints
Network efficiency and
flexibility loss recovery scheme
ve & passive ACK packore broadcast coverage
time ficiency broadcasting
ket
delivery
smission
constraints
ficiency and lif
time
ficiency
ing technique
desired reliability formation reliability
liable data fu
structure
ds upon amount of
information weight
hroughput, reduced
retransmission, delay,
energy efficiency
acket forwarding [18
nnels, dynam
logy changes and MA
overhead
and fair distribution of
energy consumption
sed on the Chinese Remaind
Theorem (CRT)
fault tolerance
ggregation and routing implemented
using beta-distribution function to
evaluate trustworthiness
Collaborative aggregation
Coefficients of a regression

L. VENKATESAN ET AL. 49
false detections in which no target is present. Hence to
r
use the measurements of a group of
gn, deployment and functional aspects of a
reliable WSN are analyzed in this paper. Also the neces-
ts of reliability in WSN application
vices provided by WSN.
REFERE
incr rac
comes necessary to f
ease the accuy of the detection algoithm it be-
sensors while discard the faulty nodes. Faults are en-
countered at any stage of WSN development, from the
high-level design of embedded software down to hard-
ware faults, energy failures, or harsh environmental con-
ditions. This makes it necessary to add fault tolerance
capability to tolerate the faulty hardware/software/fun-
ctions, so that the overall reliability of the surveillance
system can be improved. A fault diagnostic model using
case based reasoning technique in a Wireless Sensor
Network (WSN) for Wildlife Preservation application is
proposed in [42]. In the case based reasoning, semantic
tracking is applied to detect the various faults in the
hardware and software modules utilized in the sensors
over different zones, network and service levels. The
correctness and context of the data is checked using the
semantic engine with verification rules. The fault detec-
tion in WSN can be improvised using the algorithms
proposed in this work. The first part of the algorithm
decides the occurrence of fault based on a Fault Ontology
for REpetitive Semantic Tracking (FOREST) and the
second part using the Weak Inactive Link Data (WILD)
method to locate the fault in the corresponding cluster or
the device. Some of the approaches/protocol/techniques
used in WSN to enhance its reliability are compared in
Table 1.
6. Conclusion
Various desi
sity and requiremens
are discussed in detail. The existing research works with
the notion of improving reliability are studied elaborately,
compared based on their performance and reliability
measures. Many of the works are available to improve
the communication reliability of WSN by modifying the
existing MAC, routing and transport protocols. Many
methodologies to improve the sensing and coverage reli-
ability are developed exceptionally for WSN by improv-
ing the aggregation and deployment techniques. On ana-
lyzing the works, it is shown that the reliability can effi-
ciently improved by adopting fault tolerance in the de-
sign of various functions especially event/target detection
and data aggregation/fusion techniques. It is understood
that all reliability techniques concentrate in energy effi-
ciency and are aimed to improve the life time of the net-
work. At the same time less number of works are avail-
able to satisfy the other node and network constraints of
WSN and improve QoS parameters. Also need to de-
velop the concept to define the relation between security
and reliability, quantify information security, design and
testing methodologies based on the application and ser-
[1] I. F. Akyildiz, W. Su, Y. Sankarasubramaniam and E.
Cayirci, “Wireless Sensor Network: A Survey,” IEEE
Communications Magazine, Vol. 40, No. 8, 2002, pp.
102-114.
NCES
doi:10.1109/MCOM.2002.1024422
[2] J. Agre and L. Clare, “An Integrated Architecture for Co-
operative Sensing Networks,” IEEE Journal Computer,
Vol. 33, No. 510.1109/2.841788, 2000, pp. 106-108. doi:
[3] J. A. Stankovicorks,” IEEE Com- , “Wireless Sensor Netw
puter, Vol. 41, No. 10, 2008, pp. 92-95.
doi:10.1109/MC.2008.441
[4] R. Kim, J. Song and B. F. Spencer, Jr., “Reliability Analy-
sis of Wireless Sensor Networks,” Proceedings of the
1, No. 2, 2005,
, pp.
Workshop on Advanced Smart Materials and Smart
Structures Technology, Dalian, July 2011, pp. 1-12.
[5] D. Bein, V. Jolly, B. Kumar and S. Latifi, “Reliability
Modeling in Wireless Sensor Networks,” International
Journal of Information Technology, Vol. 1
pp. 1-9.
[6] H. Liu, A. Nayak and I. Stojmenović, “Fault Tolerant Al-
gorithms/Protocols in Wireless Sensor Networks” In: S. C.
Misra, I. Woungang and S. Misra, Eds., Guide to Wireless
Sensor Networks, Springer Veglag, London, 2009
261-291. doi:10.1007/978-1-84882-218-4_10
[7] W. Yuan, S. V. Krishnamurthy and S. K. Tripathi, “Im-
proving the Reliability of Event Reports in Wireless Sen-
sor Networks,” Proceedings of the 9th International
er Journal
Symposium on Computers and Communications, Vol. 1,
2004, pp. 220-225.
[8] Z. Rosberg, R. P. Liu, T. L. D. Yi, F. Dong and S. Jha,
“Statistical Reliability for Energy Efficient Data Trans-
port in Wireless Sensor Networks,” Spring
Wireless Networks, Vol. 16, No. 7, 2010, pp. 1913-1927.
doi:10.1007/s11276-009-0235-5
[9] A. Ayadi, “Energy-Efficient and Reliable Transport Pro- -
tocols for Wireless Sensor Networks: State-of-Art,”
Journal Wireless Sensor Network, Vol. 3, No. 3, 2011, pp.
106-113. doi:10.4236/wsn.2011.33011
[10] S.-J. Park, R. Vedantham, R. Sivakumar and I. F. Akyildiz,
“GARUDA: Achieving Effective Reliability for Down-
stream Communication in Wireless Sensor Networks,”
IEEE Transactions On Mobile Computing, Vol. 7, No. 2,
2008, pp. 214-230. doi:10.1109/TMC.2007.70707
[11] J. Paek and R. Govindan, “RCRT: Rate-Controlled Reli-
able Transport for Wireless Sensor Networks,” ACM
Transactions on Sensor Networks, Vol. 7, No. 3, 2010, pp.
305-317. doi:10.1145/1807048.1807049
[12] T.-W. Sung, T.-T. Wu, C.-S. Yang and Y.-M. Huang,
“Reliable Data Broadcast for Zigbee Wireless Sensor
Networks,” International Journal On Smart Sensing And
Intelligent Systems, Vol. 3, No. 3, 2010, pp. 504-520.
[13] A. G. Bagadi, S. Sarode and J. W. Bakal, “A Survey of
Reliable Transport Layer Protocols for Wireless Sensor
Network,” International Journal of Computer Applica-
tions, Vol. 33, No. 1, 2011, pp. 44-50.

L. VENKATESAN ET AL. 50
[14] K. K. Sharma, R. B. Patel and H. Singh, “A Reliable and
Energy Efficient Transport Protocol for Wireless Sensor
Networks,” International Journal of Computer Networks
& Communications, Vol. 2, No. 5, 2010, pp. 92-103.
doi:10.5121/ijcnc.2010.2506
[15] J. J. Winston and B. Paramasivan, “A Survey on Connec-
tivity Maintenance and Preserving Coverage for Wireless
Sensor Networks,” International Journal of Researc
e-
orks,” Proceedings of the
, pp. 1033-1040.
h and
Reviews in Wireless Sensor Networks, 2011, Vol. 1, No. 2,
pp. 11-18.
[16] Z. ShuKui, G. ShengRong, C. ZhiMing and F. J. Xi , “A
Reliable Collaborative Detection Scheme of Event-R
gion in Wireless Sensor Netw
International Symposium on Information Processing,
2009, pp. 021-024.
[17] G. J. Fan and S. Y. Jin, “Coverage Problem in Wireless
Sensor Network: A Survey,” Journal of Networks, Vol. 5,
No. 9, 2010
[18] G. Campobello, A. Leonardi and S. Palazzo, “Improving
Energy Saving and Reliability in Wireless Sensor Net-
works Using a Simple CRT-Based Packet-Forwarding
Solution,” EEE/ACM Transactions on Networking, Vol.
20, No. 1, 2012, pp. 191-205.
[19] E. Cañete, M. Díaz, L. Llopis and B. Rubio, “HERO: A
Hierarchical, Efficient and Reliable Routing Protocol for
Wireless Sensor and Actor Networks,” Computer Com-
munications Vol. 35, No. 11, 2012, pp. 1392-1409.
doi:10.1016/j.comcom.2012.04.003
[20] R. V. Kshirsagar and B. Jirapure, “A Survey on Fault De-
tection and Fault Tolerance in Wireless Sensor Net-
works,” International Journal of Computer Applications,
Vol. 3, No. 1, 2011, pp. 130-138.
[21] S. Mishra, L. Jena and A. Pradhan, “Fault Tolerance in
Wireless Sensor Networks,” International Journal of Ad-
vanced Research in Computer Science and Software En-
gineering, Vol. 2, No. 10, 2012, pp. 146-153.
[22] L. Paradis and Q. Han, “A Survey of Fault Management in
Wireless Sensor Networks,” Journal of Network and Sys-
tems Management, Vol. 15, No. 2, 2007, pp. 171-190.
doi:10.1007/s10922-007-9062-0
[23] F. Koushanfar, M. Potkonjak, A. Sangiovanni-Vincentelli,
“Fault Tolerance Techniques for Wireless Ad Hoc Sensor
Networks,” Proceedings of the IEEE Conference Sensors,
Vol. 2, 2002, pp. 1491-1496.
doi:10.1109/ICSENS.2002.1037343
[24] H. Sethi, D. Prasad and R. B. Patel, “EIRDA: An Energy
Efficient Interest based Reliable Data Aggregation Pro
tocol for Wireless Sensor Networ
-
ks,” International Jour-
works,” Ph.D. Thesis, The
ong Kong, 2005.
. 738-754.
a-
ternational Conference on Sensor
l
ireless Distributed Sensor Networks Subject to
nal of Computer Applications, Vol. 22, No. 7, 2011, pp.
20-25.
[25] X. Chen, “On Fault Tolerance, Performance, and Reliabili-
ty for Wireless and Sensor Net
Chinese University of Hong Kong, H
[26] E. Felemban, C.-G. Lee and E. Ekici, “MMSPEED: Multi-
path Multi-SPEED Protocol for QoS Guarantee of Reli-
ability and Timeliness in Wireless Sensor Networks,”
IEEE Transactions on Mobile Computing, Vol. 5, No. 6,
2006, pp
[27] Y.-L. Jin, H.-J. Lin, Z.-M. Zhang, Z. Zhang and X.-Y.
Zhang, “Estimating the Reliability and Lifetime of Wire-
less Sensor Network,” Proceedings of the 4th Intern
tional Conference on Wireless Communications, Net-
working and Mobile Computing, October 2008, pp. 1-4.
[28] B. L. Braem, B. Blondia, C. Moerman and I. Demeester,
“Improving Reliability in Multi-Hop Body Sensor Net-
works,” Second In
Technologies and Applications, August 2008, pp. 342-
347.
[29] J. Shin, U. Ramachandran and M. Ammar, “On Improving
the Reliability of Packet Delivery in Dense Wireless
Sensor Networks,” Proceedings of the 16th Internationa
Conference on Computer Communications and Networks,
August 2007, pp. 718-723.
[30] H. M. F. AboElFotoh, S. S. Iyengar and K. Chakrabarty,
“Computing Reliability and Message Delay for Coopera-
tive W
Random Failures,” IEEE Transactions on Reliability, Vol.
54, No. 1, 2005, pp. 145-156.
doi:10.1109/TR.2004.842540
[31] N. V. Doohan, D. K. Mishra and S. Tokekar, “Reliability
Analysis for Wireless Sensor Networks Considering En-
ansport for Wireless
Wireless Communications
vironmental Parameters Using MATLAB,” Proceedings
of the 3rd International Conference on CICSN, 2011, pp.
99-102.
[32] S. Kumar, Z. Feng, F. Hu and Y. Xiao, “E2SRT: En-
hanced Event-to-Sink Reliable Tr
Sensor Networks,” Journal of
and Mobile Computing, Vol. 9, No. 10, 2009, pp. 1301-
1311. doi:10.1002/wcm.705
[33] F. K. Shaikh, A. Khelil and N. Suri, “On Modeling the
Reliability of Data Transport in Wireless Sensor Net-
works,” Proceedings of 15th EUROMICRO International
E Pacific Rim International
sors, Vol. 12,
Conference on Parallel, Distributed and Network-Based
Processing, Feburary 2007, pp. 395-402.
[34] D. S. Kim, R. Ghosh and K. S. Trivedi, “A Hierarchical
Model for Reliability Analysis of Sensor Networks,”
Proceedings of the 16th IEE
Symposium on Dependable Computing, December 2010,
pp. 247-248.
[35] I. Silva, L. A. Guedes, P. Portugal and F. Vasques, “Reli-
ability and Availability Evaluation of Wireless Sensor
Networks for Industrial Applications,” Sen
No. 1, 2012, pp. 806-838. doi:10.3390/s120100806
[36] L. Xing and H. E. Michel, “Integrated Modeling for Wire-
less Sensor Networks Reliability and Security,” Proceed-
ings of the Annual Conference on Reliability and Main-
tainability Symposium, January 2006, pp. 594-600.
, pp.
ol. 5,
[37] Y.-F. Xiao, S.-Z. Chen, X. Li and Y.-H. Li, “Reliability
Evaluation of Wireless Sensor Networks Using an En-
hanced OBDD Algorithm,” Proceedings of the IEEE In-
ternational Conference on Communications, 2009
453-457.
[38] W. W. Bein, D. Bein and S. Malladi, “Reliability and
Fault Tolerance of Coverage Models for Sensor Net-
works,” International Journal of Sensor Networks, V
No. 4, 2009, pp. 199-209.
doi:10.1504/IJSNET.2009.027628
[39] V. Kumar, R. B. Patel, M. Singh and R. Vaid, “Reliability

L. VENKATESAN ET AL. 51
-79.
ust 2011, pp. 1-6.
as, “Data Fusion with
.2010.93
Analysis in Wireless Sensor Networks,” International
Journal of Engineering and Technology, Vol. 3, No. 2,
2011, pp. 74
[40] A. Munir and A. Gordon-Ross, “Markov Modeling of
Fault-Tolerant Wireless Sensor Networks,” Proceedings
of 20th International Conference on Computer Commu-
nications and Networks, Aug
[41] H. Luo, H. Tao, H. Ma and S. K. D
Desired Reliability in Wireless Sensor Networks,” IEEE
Transactions on Parallel and Distributed Systems, Vol.
22, No. 3, 2011, pp. 501-512.
doi:10.1109/TPDS
nications
, Sensor Net-
ssing, 2011, pp. 377-382.
ettergren, “Coverage and Reli-
[42] V. Latha, C. Subramaniam and S. Shanmugavel, “Fault
Tolerant Wireless Sensor Network Using Case Based
Reasoning with Semantic Tracking,” Proceedings of the
WSEAS International Conference on Commu,
July 2011, pp. 240-246.
[43] M. A. Mahmood and W. K. G. Seah, “Event Reliability in
Wireless Sensor Networks,” Proceedings of 7th Interna-
tional Conference on Intelligent Sensors
works and Information Proce
[44] K. El-Darymli, F. Khan and M. H. Ahmed, “Reliability
Modeling of Wireless Sensor Network for Oil and Gas
Pipelines Monitoring,” Sensors & Transducers Journal,
Vol. 106, No. 7, 2009, pp. 16-26.
[45] E. G. Rowe and T. A. W
ability of Randomly Distributed Sensor Systems with
Heterogeneous Detection Range,” International Journal
of Distributed Sensor Networks, Vol. 5, No. 4, 2009, pp.
303-320. doi:10.1080/15501320802299853
[46] A. Taherkordi, M. A. Taleghan and M. Sharifi, “Achieving
Availability and Reliability in Wireless Sensor Networks
Applications,” Proceedings of the 1st International Con-
ference on Availability, Reliability and Security, 2006, pp.
529-535. doi:10.1109/ARES.2006.21
[47] A. Zaidi, B. O. Bouamama and M. Tagina, “Bayesian
Reliability Models of Weibull Systems: State of Art,” In-
ternational Journal of Applied Mathematics and Com-
puter Science, Vol. 22, No. 3, 2012, pp. 585-600.
and R. Fonseca, “Reliable Transfer on Wireless
ari and L. U. Kadam, “Wireless Sensor Net-
n by Adaptive Network Coding in Wireless Sensor
te-
[48] P. S. Patheja, A. Waoo and P. Shrivastava, “Fault Tolerant,
Energy Saving Method for Reliable Information Propaga-
tion in Sensor Network,” International Journal of Smart
Sensors and Ad Hoc Networks, Vol. 1, No. 4, 2012, pp.
34-38.
[49] L. Venkatesan, C. Subramaniam and S. Shanmugavel,
“Reliability Modeling of Context Aware Wireless Sensor
Network,” International Journal of Information and Ele-
ctronics Engineering, Vol. 2, No. 5, 2012, pp. 710-715.
[50] S. Kim
Sensor Networks,” Proceedings of the 1st Annual IEEE
Communications Society Conference on Sensor and Ad
Hoc Communications and Networks, October 2004, pp.
449-459.
[51] Z. S. Bojkovic, B. M. Bakmaz and M. R. Bakmaz, “Secu-
rity Issues in Wireless Sensor Networks,” International
Journal of Communications, Vol. 2, No. 1, 2008, pp. 106-
115.
[52] H. C. Chaudh
works: Security, Attacks and Challenges,” International
Journal of Networking, Vol. 1, No. 1, 2011, pp. 04-16.
[53] T.-G. Li, C.-C. Hsu and C.-F. Chou, “On Reliable Trans-
missio
Networks,” Proceedings of the IEEE International Con-
ference on Communications, June 2009, pp. 1-5.
[54] H. Kwon, T. H. Kim and S. Choi, “A Cross-Layer Stra
gy for Energy-Efficient Reliable Delivery in Wireless
Sensor Networks,” IEEE Transactions on Wireless Com-
munications, Vol. 5, No. 12, 2006, pp. 1-11.
doi:10.1109/TWC.2006.256992
[55] S. Saqaeeyan and M. Roshanzadeh, “Improved Multi-Path
and Multi-Speed Routing Protocol in Wireless Sensor
Networks,” International Journal of Computer Network
and Information Security, Vol. 4, No. 2, 2012, pp. 8-14.
doi:10.5815/ijcnis.2012.02.02
[56] T. Ryutov and C. Neuman, “Trust based Approach for Im-
proving Data Reliability in Industrial Sensor Networks,”
ftp://ftp.isi.edu/isi-pubs/tr-631.pdf
[57] K. Maraiya, K. Kant and N. Gupta, “Efficient Cluster
Head Selection Scheme for Da ta Aggregation in Wireless
Sensor Network,” International Journal of Computer Ap-
plications, Vol. 23, No. 9, 2011, pp. 10-18.