A Survey on Topology Control and Maintenance in Wireless Sensor Networks

Indonesian Journal of Electrical Engineering and Informatics (IJEEI)
Vol. 2, No. 2, June 2014, pp. 96~100
ISSN: 2089-3272  96
Received March 16, 2014; Revised May 12, 2014; Accepted May 25, 2014
A Survey on Topology Control and Maintenance in
Wireless Sensor Networks
Manish Singh
Department of Electronics and Communication Engineering
Invertis University, Bareilly, (UP) INDIA
e-mail: manishsingh.msingh@gmail.com
Abstract
Wireless Sensor Networks (WSNs) consist of devices equipped with radio transceivers that
cooperate to form and maintain a fully connected network of sensor nodes. WSNs do not have a fixed
infrastructure and do not use centralized methods for organization. This flexibility enables them to be used
whenever a fixed infrastructure is unfeasible or inconvenient, hence making them attractive for numerous
applications ranging from military, civil, industrial or health. Because of their unique structure, and limited
energy storage, computational and memory resources, many of the existing protocols and algorithms
designed for wired or wireless ad hoc networks cannot be directly used in WSNs. Beside this, they offer a
flexible low cost solution to the problem of event monitoring, especially in places with limited accessibility
or that represent danger to humans. Applications of large scale WSNs are becoming a reality example are
being a Smart Grid, Machine to Machine communication networks and smart environment. It is expected
that a topology control techniques will play an important role in managing the complexity of such highly
complicated and distributed systems through self-organization capabilities. WSNs are made of resource
constrained wireless devices, which require energy efficient mechanisms, algorithm/protocol. Control on
topology is very important for efficient utilization of networks and is composed of two mechanisms,
Topology Construction (TC) and Topology Maintenance (TM). By using these mechanism various
protocols/algorithm have came into existence, like: A3, A3-Coverage (A3-Cov), Simple Tree, Just Tree,
etc. This paper provides a full view of the studies of above mentioned algorithms and also provides an
analysis of their merits and demerits.
Keywords: WSNs, A3, A3-Cov, Simple Tree, Just Tree
1. Introduction
WSNs have become an emerging technology that has a wide range of potential
applications including environment monitoring, object tracking, scientific observing and
forecasting, traffic control, etc [1] [2]. It normally consists of a large number of distributed nodes
that organize themselves into a multi-hop wireless network and typically these nodes co-
ordinate to perform a common task [3].
For efficient use of WSNs i.e. it functions properly its topology should be control and
maintain time to time. For the first method, TC mainly focuses on constructing a connected
topology while minimizing energy consumption of nodes to extend the network lifetime. Second
i.e. TM it maintains the existing topology, when the existing can’t provide the requested service
any longer.
Topology construction is important to control the required topology. It is an important
technique used in WSNs to achieve energy conservation and extend network lifetime without
affecting important network performance such as connectivity and throughput [6]. Topology
provides information about a set of nodes and connectivity (links) between a pair of nodes in the
set. To construct a network topology, each sensor node discovers its neighbors and relative
links using its maximum transmission power. Based on the information gathered the node can
make decisions to build a network.
The rest of this paper is organized as follows: Section 2, introduces the related work.
Section 3, compares merits and de-merits between A3 and A3-Cov, Simple Tree and Just Tree.
Section 4, concludes this paper and also highlights future scope of the work.

 ISSN: 2089-3272
IJEEI Vol. 2, No. 2, June 2014 : 96 – 100
97
2. Related Work
Apart from constructing a topology by reducing the transmission range, other
techniques is also present based on the assumption that nodes have information about their
own positions and the position of their neighbors or they have directional antennas that are
used to determine the orientation of the nodes.
Other method is based on the Connected Dominating Set (CDS) [1] paradigm. The idea
is not to change the transmission range of the nodes but to turn unnecessary nodes off while
preserving important network properties, such as connectivity and communication coverage.
Based on above mentioned topology construction methods some algorithms are discussed in
this section.
2.1. A3
The A3 algorithm [4] [5] produces an approximate solution to the minimal CDS [1]
problem. The A3 algorithm assumes no prior knowledge about the position or orientation of the
nodes; therefore, the nodes do not have an exact geometric view of the topology. However,
nodes can determine how far a node is based on the strength of the signal received, and this
information is enough to select a close-to-optimal CDS tree, based on the belief that farther
nodes will offer better area of communication coverage. The A3 algorithm is executed in 2
moments: Neighborhood discovery, children selection.
All nodes start with the unvisited state, except the stating node, which starts with the
“Active” candidate state. An active candidate node sends a “Hello Message” to all its
neighbours. The first one that sends this message is the sink node. In addition, this node sets a
timer to wait for replies from unvisited neighbour nodes. All the neighbors send back a “Parent
Recognition” message that includes their ID and their own selection metric, which is a convex
combination of the ratio or remaining energy in the receiver, and the ratio of distance over the
maximum transmission range. Also they adopt the sender as their “Parent nodes” and change
their state to child.
After a period of time, the active candidate node stops listening for messages, sorts the
list of “children” nodes (neighbors who answered) in a decreasing order, and sends this sorted
list back to its children. If the active candidate node has received at least one answer, it will
change its state to active; otherwise, it will change its state to “sleeping” and will turn off its
components until the next topology maintenance routine is executed.
The children nodes find themselves in the list and wait for a period of time proportional
to their position on the list. When the timer in a node expires, and it has not received any
“Sleeping” messages, the node will send a “Sleeping” message, change its state to active
candidate and if the node receives a “Sleeping” message while in the timer set, it will change its
state to “Sleeping Candidate”, and will turn off its component for a period of time. After this timer
expires, the node will change its state to active candidate.
2.2. A3-Cov
A3-Cov algorithm works [4] very similar to A3 protocol, but presents important changes
in some portion: like, if there are any nodes that have not received any “Parent Recognition”
message, it means that there are no nodes that depend on it for communication purposes;
however, they may still be useful in order to extend the network’s sensing coverage. In order to
do this, A3-Cov defines a new variable in the nodes called “sensing covered” i.e. node “x” is
sensing covered by node “y” if “x” is inside the sensing range of “y’ and “y” is an active node.
In A3 algorithm after the timer expire in node to receive “Sleep” message. If the node
has been “Sensing Covered” by any other node (including its parent node), it sets a short timer
to wait for “Sensing Covered” message from its active neighbor.
If the timer expires and the node is not “Sensing Covered” yet, it will turn itself on,
changes its state to active and send a “Sensing Covered” message and a “Sleeping” message.
If any node in its range receives the “Sensing Covered” message, it will evaluate if it has been
covered by sender, in which case it will update the value of the “Sensing Covered” variable.

IJEEI ISSN: 2089-3272 
A Survey on Topology Control and Maintenance in Wireless Sensor Networks (Manish Singh)
98
If the node received a “Sensing Covered” message from any other node, it will stop the
timer changes its state to “Sleeping” and turn its component off until the next topology
maintenance routine.
A3Cov expands considerably the coverage area as compare to A3.
2.3. Just Tree
For the homogeneous network number of nodes, the deployment area, sink node and
the Virtual Network Interface (VNI) [7], play an important role. The just tree algorithm assumes
one sink node responsible for message/ information broadcast. The sink nodes are capable of
sending or receiving messages from other neighboring sensor nodes. If CDS rule-k is applied
and the topology is constructed, this CDS rule-k is needed to run for a quite number of times, a
lot of energy will be spent to maintain a particular topology if a sink node prefers to broadcast or
send. The concept of spanning tree is considered in most of the cases. This concept can be
employed for number of nodes starting from 50-1000 nodes or even more, but in order to
achieve acceptable results the node number is increased in multiple of 100. As far as just tree
protocol/algorithm is considered, less energy is spent with a reduced queue size, if number of
nodes is increased. The message or number of events are propagated within the network using
the same concept of parent node and child node, the parent node initiates the message and
transfer this message to other sensing nodes acting as child node. Keeping itself in a dominant
position i.e. if CDS rule-k is taken into consideration the parent node has maximum in build
energy which gradually reduces as the number of nodes increases and message transfer takes
place i.e. if the size of the tree is large (more number of nodes) the total energy spent will
ultimately decreases till the last child node is covered in a general prospective, if the tree is
giant it means that it will cover a larger deployment area and will have large number of children
nodes connected with the parent node, which will be the main source of energy with some
threshold value i.e. such topology will require atleast.
(i) Homogenous network
(ii) Flexible deployment area
(iii) Parent node that initiates a “HELLO” message with same reasonable amount of threshold
in terms of energy in order to support varying Queue size if number of children nodes are
also varied.
(iv) The recognition of the initiated “HELLO” message must be acknowledge by children nodes
in order to estimate queue size, energy consumption, number of messages transfer during
simulation, performed for different number of nodes for different time periods. The concept
of just tree ensures that as the deployment area will increases or if the deployment area is
constant the number of nodes if increased will denote the increase in the size of the tree in
order to efficiently cover a flexible or constant deployment area.
2.4 Simple Tree
Simple Tree is a derivative form of one or more derivative of spanning tree derived from
the just tree algorithm which considers only one CDS per one just tree. According to this, if this
algorithm is further splitted into more than one CDS rule-k [1] the load on single parent node can
be slightly reduced, however, the total energy spent may substantially increased and it may also
affect the queue size. But such algorithm will also require more simulation time because there
exist more number of subsets in the same deployment area for the increased node density i.e.
such algorithm are not that much simple as far as their name highlights, but are complex
requires a greater degree of simulation efforts, are hard to model, are dependent on large
queue size and lastly at the out-set shows high value of energy i.e. spent energy. The only
advantage of these type of algorithm lies in the aspect that they ensure complete message
distribution within their individual CDS and further, if area of the topology is very sparse it can
easily recovered by introducing a new CDS in the form of a simple tree rather than raising or
disturbing the pre-existing just tree that have their individual CDS. Simple tree algorithm can
also be modified to operate for heterogeneous network, if the CDS functions as a closed loop
and even if the topology is homogenous and if the CDS functions as a close loop the number of

 ISSN: 2089-3272
IJEEI Vol. 2, No. 2, June 2014 : 96 – 100
99
event or messages floated within the complete network will be less as compared to the just tree
algorithm.
3. Comparative Analysis
Based on above discussed algorithm, by analyzing them, some important observations
can be make-out between them. The important observations in support of the above discussed
algorithm are listed in Table 1 and Table 2 respectively.
Table 1. Comparison between A3 and A3-Cov
Algorithm Merits Demerits
A3
1. A3 does not need location information: no GPS
or any Location mechanism is necessary.
2. A3 is simple and present Low computational
complexity.
3. A3 requires no synchronization scheme.
4. No effect on topology, as if number of nodes
increases its response time is almost same.
1. Number of message or data transfer
rate is low in A3 protocol.
2. Coverage area in A3 is Less compared
to A3-Cov protocol.
A3-Cov
1. Most of the merits are same as like A3 protocol
except:
2. Coverage area is more than A3protocol.
3. Data transfer rate is more than A3 protocol.
1. As number of nodes increases in A3-
Cov response time varies acc. to nodes.
2. Energy consumption is more compared
to A3 protocol
Table 2. Comparison between Simple Tree and Just Tree
Algorithm Merits Demerits
Simple Tree
1. Applicable for heterogeneous as well as
homogeneous close loop networks.
2. Complete distribution of message takes
place.
3. Covers larger deployment area.
4. Energy spent on individual node
increases.
1. Difficult to simulate.
2. Network complexity increases
(as more than one just tree exist).
3. Response time is high.
Just Tree
1. Reduced queue size compared to simple
tree.
2. Energy spent is less as compared to
simple tree.
3. Response time is low as compared to
simple tree.
1. Covers less area as compare to
simple tree.
2. Message broadcast depends
entirely on the single parent node.
3. Suits only homogeneous
networks.
4. Conclusion
The affect of existing algorithm for topology maintenance and topology construction
have been discussed in this paper. This paper addresses: A3, A3-Cov, Simple Tree, Just Tree
and provide an initial review of their performance in TM and TC for sparsely deployed WSNs.
The work presented in this paper is a basic attempt towards analyzing and highlighting the
performance metrics of the discussed algorithm.
In future these individual algorithms can be simulated to have a deeper insight for
further optimizing their utility in varied type of WSNs.
References
[1] A Karthikeyan, T Shankar, V Srividhya, Siva Charan Reddy V, Sandeep Kommineni. “Topology
Control Algorithm for Better Sensing Coverage with Connectivity in Wireless Sensor Networks”.
Journal of Theoretical and Applied Information Technology. 2013; 52(3).
[2] Degan Zhang, Guang Li, Ke Zheng, Xuechao Ming and Zhao-Hua Pan. “An Energy-Balanced Routing
Method Based on Forward-Aware Factor for Wireless Sensor Networks”. IEEE Transactions on
Industrial Informatics. 2014; 10(1).
[3] Mo Li and Baijian Yang. “A Survey on Topology issues in Wireless Sensor Network”.

IJEEI ISSN: 2089-3272 
A Survey on Topology Control and Maintenance in Wireless Sensor Networks (Manish Singh)
100
[4] Pedro Mario Wightman R, Miguel A. Labrador. “Reducing the communication range or turning nodes
off? An initial evaluation of topology control strategies for wireless sensor networks”. Ingeniería &
Desarrollo. Universidad del Norte. 2010; 28: 66-88.
[5] Pedro M. Wightman and Miguel A. Labrador. “A3: A Topology Construction Algorithm for Wireless
Sensor Networks”. IEEE. 2008.
[6] Azrina Abd Aziz, Y Ahmet S¸ekercio˘glu, Paul Fitzpatrick, and Milosh Ivanovich. “A Survey on
Distributed Topology Control Techniques for Extending the Lifetime of Battery Powered Wireless
Sensor Networks”. IEEE Communications Surveys & Tutorials. 2013;15(1).
[7] Pedro M Wightman, Pedro M Wightman. “Atarraya: A Simulation Tool to Teach and Research
Topology Control Algorithms for Wireless Sensor Networks”. ACM ISBN 978-963-9799-45-5,
Simulation tool. 2009

A Survey on Topology Control and Maintenance in Wireless Sensor Networks

More Related Content

What's hot (16)

Similar to A Survey on Topology Control and Maintenance in Wireless Sensor Networks (20)

More from ijeei-iaes (20)

Recently uploaded (20)

A Survey on Topology Control and Maintenance in Wireless Sensor Networks