Research A rticle
Dynamic Clustering Algorithm for Tracking Targets with
High and Variable Celerity (ATHVC)
Mohamed Toumi, Abderrahim Maizate, Mohammed Ouzzif, and Med Said Salah
RITM-ESTC/CED-ENSEM, University Hassan II, Casablanca, Morocco
Correspondence should be addressed to Mohamed Toumi; m
toumy@yahoo.fr
Received July ; Accepted October
Academic Editor: Yun Liu
Copyright © Mohamed Toumi et al. is is an open access article distributed under the Creative Commons Attribution
License, which permits unrestricted use, distribution, and reproduction in any medium, provided the original work is properly
cited.
Target tracking with the wireless sensors networks is to detect and locate a target on its entire path through a region of interest.
is application arouses interest in the world of research for its many elds of use. Wireless sensor networks, thanks to their
versatility, can be used in many hostile and inaccessible to humans environments. However, with a limited energy, they cannot
remain permanently active, which can signicantly reduce their lifetime. e formation of a cluster network seems an eective
mechanism to increase network lifetime. We propose to build optimal dynamic clusters on the target trajectory. For increasing
energy eciency, our algorithm integrates for the rst time, to our knowledge, strategies to avoid overlapping clusters and a model
to wake up the s ensors, adapting to the context of targets with large and variable speed.
1. Introduction
ere are many areas of use of the wireless sensor networks;
we can mention military applications, environmental and
industrial monitoring, applications related to smart cities,
and others. We are interested in monitoring applications and
in particular the tracking of moving targets. Target tracking
is to perform two main functions: the detection and the mon-
itoring of the target along its path through sensors deployed
in an area of interest. However, the deployment of these net-
works depends on the constraints imposed by the miniatur-
ized structure of the sensors: small storage capacity, limited
life of the battery, communication range, bandwidth, and so
forth. Particularly, as in most cases, the nodes are not ser-
viceable aer their deployment. Forming a clustered network
seems an eective mechanism to increase network’s lifetime.
Many researchers focused on the development of ecient
energy clustering algorithms [–]; however, these algorithms
are not suitable for applications specications related to the
targets tracking. Indeed, the cited protocols oer clustering
schemes that form and maintain a hierarchical network
without considering the fact that the target itself seems rare
and in specic places in the network. e optimization of the
clustering algorithms consists in activating only the nodes
which are in the path of the target when it is within their
detection ranges. All other sensors must be in sleep mode.
Numerous proposals have been published in recent years [–
] which propose dynamic clustering algorithms; these form
temporal clusters depending on the evolution of the target
across the network. However, despite the eciency of their
energy,thesealgorithmsarenotadaptedtoanenvironment
where the velocity of the target can become extremely large
and variable. Tracking, for example, the tsunami waves whose
velocity ≈870
(∗6)is a function of the single parameter
water depth (km) in the case of suciently long period
of tsunamis, typically about ten minutes, which is the case
of most tsunamis of tectonic origin. is means that the
speed is km/h for a depth of km and km/h for a
depth of m. e variability of the speed of a tsunami is
clearly identiable with the approach of the coast. Tracking
wave of an earthquake or a wildre driven by the torrential
winds are examples of applications where speed of the target
is important and variable. In this case, the target can be
moved, while the cluster conguration’s messages are just
in the neighbor discovery phase and sharing information.
e time required for the selection of cluster heads, based
Hindawi Publishing Corporation
Journal of Computer Networks and Communications
Volume 2016, Article ID 7631548, 10 pages
http://dx.doi.org/10.1155/2016/7631548
Journal of Computer N etworks and Communications
on dierent metrics and recruitment of member nodes, can
be detrimental to ensure the function of tracking for this
type of applications. We propose an ecient algorithm that
is well suited to target tracking applications with sig nicant
and variable speed; in our algorithm, we use a new metric
for forming optimal clusters according to the evolution of the
target, with minimal time waiting and clusters overlap, which
allows for better energy eciency and especially having a
minimum response time needed to track high-speed targets.
e selection of cluster head is not the only problem to
be raised in the dynamic clustering algorithms. Indeed, the
nodes are a priori in the sleep state; only the sensors that are
onthetargetpathmustbeactive.Sohowandbywhatcriteria
should we activate these sensors to form the active cluster to
retrieve the data on the target to a base station?
e prediction schemes have b een proposed in recent
years to predict the position of the target which enables
activating only the nodes which are on the trajectory of
the target. e extended Kalman lter [] combined with
detection mechanisms for changes of direction as CuSum []
can eectively calculate future coordinates of the target and
wake up the sensors accordingly. e prediction lters require
message exchanges and sometimes complex calculations by a
central entity, which consumes a lot of time, particularly as
more oen these algorithms use a correction step and lead
to additional calculations and thus a precious time is wasted
which is necessary to eectively meet the real time constraint
imposed by high-speed target. Our prediction system must
necessarily take account of this constraint. We propose to
use an activation based on overhearing the members of the
active cluster to wake the other nodes on the way to the target.
e sensors that are awakened must return to the sleep state
once they no longer detect the target aer a system time .
So the question is how to properly adjust this time in the
context of variable speed targets. Poor timing could create
what is called “the reected wake waves.” Although several
solutions using this method of activation exist in the literature
[], our solution is, to our knowledge, the rst to consider a
prediction scheme oering two eective solutions to prevent
the problem of reected wake waves descr ibed in Section
and to minimize the overlap of the generated clusters during
the movement of the target.
is article is organized as follows: in Section , we
present some related work on tracking of the targets. en, in
Section , we will detail the proposed protocol. In Section ,
we present the simulation results. Finally, we end this article
with a conclusion in Section .
2. Related Work
An important number of researches concerning tracking
moving targets in the context of wireless sensor networks
with the use of clustered network architectures exist in the
literature. is type of architecture provides advantages such
as scalability, trac reduction, and energy eciency. ese
works can be classied into three main categories:
(i) Static clustered architectures
(ii) Dynamic clustered architectures
(iii) Hybrid clustered architectures
2.1. Static Clusters. Algorithms [–] are trying to statically
create clusters at the time of the network deployment based
on dierent criteria for the CH (cluster head) selection, such
as the position, density, security, computing power, or the
residual energy of the sensors. However, these static cluster-
ing algorithms are not suitable for targets tracking, be cause
they hold all the nodes in activities, even if the target appears
only rarely. It is inecient to create groups of the nodes in
advance regardless of the target’s movement in the sensor
eld; this can waste ineciently the sensors’ energy.
2.2. e Dynamic Clusters. In this category of algorithms, the
target’s trajectory is predicted, allowing only the activation of
the sensors in the path and thus saving energy. is prediction
can be performed using predictive models including the
Kalman lters [] or using probabilistic mechanisms such as
Markov chains [].
e authors of [] presented the prediction protocol and
sleep scheduling nodes based on the probability (SSPP) to
imp rove energy eciency. e approach provides a target
prediction method based on kinematics and probability.
An approach is proposed in [] to awaken the sensors
that form clusters along the planned trajectory to reduce
the probability of missing the target. e solution provides
that active cluster members identify the target and send the
data to their cluster head. e base station collects all the
data members, determines the possible location of the target,
and sends wake-up messages to the nodes in the path of the
target.
In [], there is a provided dynamic clustering algorithm
coupled with the Kalman lter to predict the position of
the single moving target. e Kalman lter is a prediction
model with two stages: prediction and correction. It allows
estimating recursively the process status based on its earlier
statements; it aims to estimate the future target position based
on the current position. It is described using a state evolution
model and a measurement model that assumes linear with
Gaussian errors.
e authors of [] propose a distributed algorithm for
constructing clusters dynamically and measuring the dis-
placement of the target. Sensors that detect the target enter a
regional competition regime. Literally, every such candidate
must calculate a parameter called CHEW. e node with the
lowest value will start the operation of the recruitment of
member’s knots. e equation of CHEW parameter is dened
by
CHEW =INT
target
batry
+random().
()
e autho rs want to promo te the node with a high resid-
ualenergyandthatclosesttothetarget.esecondpart
decimal random() is a random value between and whose
role is to prevent the collision of data transmission when two
or more nodes dene the same window size.
CHEW creates an optimal clustering scheme depending
onthetargetofthemovementasshowninFigureandselects
from its neighboring sensors that participate in the monitor-
ing of the target. e predictive solutions have the advantage
Journal of Computer Networks and Communications
F : Diagram clustering of CHEW.
to exploit with the best way the available information to
save energy by activating only nodes that are in the path of
the target. However, they remain unsuitable to the contexts
of the target with very large and variable speeds. Indeed,
the predictions calculation time can be disadvantageous to
respond with a real time for tracking applications associated
with a great target speed.
2.3. Hybrid Solution. As its name suggests, this category
includes solutions with several combined approaches. For
example, in [, ], the authors propose algorithms hybrid
clustering in which dynamic reactive clusters are formed in
collaboration with static proactive clusters. e major objec-
tive of this research is to continue to monitor targets in the
cluster border regions with energy eciency; ho wever, these
algorithms still suer from energy ineciency due to having
a priori static cluster structure.
In our work, we have oered a global vision by focusing
on collaboration between nodes, including how to wake them
up one aer the other to follow the target throughout its
evolution in the area of inter est, and managing at best the
energy consumption by minimizing the overlap between the
clusters and avoiding the problem of the reected wake waves;
then we have proposed a metric for selecting clusters head
adapting to the context of large and variable sp eed targets.
e next paragraphs will be applied to describe in detail this
solution.
3. The Proposed Algorithm
3.1. e System M odel and the Assumptions. We assume that
nodes are initially in the state of sleep which guarantees
minimal energy consumption. Indeed, in this state, all equip-
ment units, which make up the sensor , are o, except for a
processing unit and a low power channel for receiving the
activation-up messages. Upon receiving a wake-up message,
each node has to start all these hardware units.
It is also assumed that the nodes have knowledge of their
geographical positions and the rst target detection is done.
istaskisbeyondthescopeofthisproject.
In general, the signal received by node
𝑖
from node
𝑗
is decreased with the distance between the two nodes.
We are adopting the mitigated disk detection model []
to estimate the distance (
𝑖
,
𝑗
)from the received signal.
e model equation is written as follows:
𝑖
=
𝛼
𝑖
,
𝑗
if
𝑖
,
𝑗
≤
𝑠
0 else.
()
𝑖
isthesignalreceivedfromthenode
𝑗
. is the original
strength of the signal transmitted by a node. is attenuation
coecient depending on the environment.
𝑠
is the detection
range of the node. (
𝑖
,
𝑗
) is the Euclidean distance
between t he two nodes
𝑖
and
𝑗
.
To facilitate the description of the protocol, we adopt the
following notations:
(i)
𝑡
is node communication range.
(ii)
𝑠
is the node detection radius.
(iii) (
𝑖
,
𝑠
) is the detection region of node
𝑖
with the
detection range
𝑠
.
(iv) is the wireless sensor network G =(V,E).
(v) is the set of all links between nodes. e set is
dened by ==(V
𝑖
,V
𝑗
)|{V
𝑖
,V
𝑗
∈
2
∩(V
𝑖
,V
𝑗
)≤
𝑡
∩ =}.
(vi) V is the set of all nodes: V ={v
1
,v
2
,...,v
n
}.
(vii) e cluster C
i
is dened by C
i
:{v
j
| d(v
i
,v
j
)≤R
t
};
v
i
is the cluster head node.
(viii) L
(t)
isthelocationofthetargetattime.
3.2. e Activation of Nodes in Sleep State (Prediction). ere
are two types of targets in relation to their ability to commu-
nicate with the network: targets that can communicate and
targets that lack this ability.
A target that can communicate is a target equipped with
a communication module allowing it to transmit signals or
periodic messages (hello messages, for example). On the
other hand, the no-communicating targets are devoid of this
capacity; they are more realistic especially when it comes to
model natural phenomena. We are interested in this kind of
target in the following: the detection process becomes more
complexasthesetargetsarenotcooperative.Inourcon-
tribution, we propose an algorithm that uses the activation
of sensors located on the way to the target by overlistening
nodes belonging to the active cluster. e activation process
is described as follows.
Let
𝐾
:{V
𝑗
,(CH,V
𝑗
)≤
𝑡
}be the set of the CH (cluster
head)nodeneighborhoodsoftheactivecluster,whichis
dened as
Neig
=V
𝑖
∈∩
𝐾
|∃V
𝑗
∈
𝐾
,V
𝑖
,V
𝑗
≤
𝑡
. ()
Neig
is the set of all neighboring nodes to the active cluster
𝐾
.
𝐾
={V
𝑗
∈
𝐾
|∃
0
,(
(𝑡
0
)
,V
𝑗
)≤
𝑠
} is the set of
the nodes belonging to the active cluster which detected the
target at some point. It also denes the set of the nodes that
Journal of Computer N etworks and Communications
Target
K
B
CH
R
t
r
s
R
t
A
R
t
B
A
F : Activation of nodes in sleep state.
will be activated by the prediction process based on three
overhearing messages by
Activ
=V
𝑖
∈∩
𝐾
|∃V
𝑗
∈
𝐾
,V
𝑖
,V
𝑗
≤
𝑡
. ()
It is clear that
Activ
Neig
;thatistosay,theset
of the nodes that will be activated is only a subset of the
active cluster neighbors. is improves energy eciency by
activating just the nodes that potentially will detec t the target.
In Figure , the active cluster is dened by the transmis-
sion range
𝑡
of the CH. All nodes of the chopped area will
move to the active state; nodes and which have detected
the target and which are at the edge of the active cluster mark
thewakezonebytheirrespectivetransmissionrange:
𝑡
and
𝑡
.
3.3. Clusters O verlap Problem. e majority of the dynamic
clustering algorithms suer f rom clusters overlap problem.
Figure (a) is showing that cluster is still operational because
the target is in range of node , while another cluster will be
formed when node detectsthetargetatpoint.
To minimize the overlap time, we dene an area where
nodes, awakened by the overhearing messages, will have a
status CH
Forb;thatistosay,theycannotbeCH,evenifthey
are the rst to detect the target. is eld is the set dened by
N
Forb
=v
i
∈N
Activ
|∃v
j
∈D
k
,v
i
,v
j
≤
s
. ()
In Figure (b), the nodes belonging to the chopped area
dened by the detection ranges
𝑠
and
𝑠
,respectively,of
nodes and will not participate in the selection of the
cluster head; they will have necessarily a member status.
It therefore sets a requirement that node V will have the
status Cluster
Ready;thatistosay,itcouldbeaercluster
head.
e set of the neighbors that detected the target of node
k belonging to the active cluster is dened by
N
Node-Act
(
k
)
=
v
i
∈D
k
|
v
i
,v
≤R
t
.
()
Node V is CH
Ready if and only if
V k
i
∈N
Node-Act
(
k
)
min k
i
,k>
𝑠
.
()
is simply can be t ranslated by imposing the following
condition.
Each node thatisnotmemberoftheactivecluster
receiving a message Msg-data from a member node k
i
must
evaluate the distance (k
i
,k) separating the two nodes, if the
condition (k
i
,k)≤
𝑠
is veried; node V cannot be CH and
passes to CH
forb status.
3.4. e Selection of CHs. e target tracking phase is trig-
gered by building optimal clusters aer the target has been
detected. e cluster-building process is illustrated in Fig-
ure . Recall that a node cannot become a CH unless it has the
status CH
Ready; in other words the node belongs to N
ready
group which is dened as
N
ready
=v
i
∈N
Activ
|Vv
j
∈D
k
,v
i
,v
j
>
𝑠
. ()
If this is no t the case, the node can only become a member
pending an invitation from a CH or returns to the sleep state
aer a system time.
Each node V having the status CH
Ready and detecting
the target must generate timer competition aer which it
passes to the node cluster head state and subsequently send
the messages MSG
Inv to recruit member’s nodes.
is comprised of two parts, as shown in (). e rst
term of aims to create repulsion eect between adjacent
clusters and therefore allows distances between clusters one
another.
Indeed, the more the candidate node to be CH is far
from the active cluster nodes, the more it will have a small
competition timer and therefore a greater probability of being
a cluster head. is metric reduces overlap and at the same
time the number of generated cluster s and therefore better
energy eciency:
=
𝑠
min
(
V
)
+random(),
()
min
(
V
)
=min V
𝑖
,V, V
𝑗
∈
Node-Act
(
V
)
. ()
e second term of is a random number between
and ; its role is to prevent two or more candidates nodes
generating the same .
It is easily shown that < 2ms, which is not a
considerabletime,becausethetargetwouldnothavelethe
scope of detecting even with very high speeds. Figure shows
the distance traveled by a target at dierent speeds in a
ms time. At a speed of km/h, for example, the target
would have traveled . meters against meters at a speed of
, km/h.
e proposed solution reduces the number of clusters
formed during the displacement of the target. Indeed, without
the new proposed metric, we form more compact clusters
with a considerable overlap as shown in Figure (a). Our
Journal of Computer Networks and Communications
Targe t
K
A
B
K
A
B
CH
R
t
r
s
(a)
Targe t
K
A
B
CH
R
t
A
R
t
B
r
s
A
r
s
B
N2
(b)
F : Overlap problem between clusters.
0
5
10
15
20
10000 20000 30000 40000
0
e target speed (km/h)
e distance traveled (m)
F : e t raveled distance in ms.
solution is trying to acquire a scope increase in each cluster
so as to minimize the overlap as shown in Figure (b).
Fewer clusters will be formed to the same movement of
the target, which will necessarily save total energy consump-
tion.
3.5. e Problem of Reected Wake Waves. When the target
moves through the active cluster (cluster in the diagram),
the nodes in this cluster use data messages (Msg
Data) to
collect and send the data to their cluster head, waking at
thesametimethenodesintheoutsideofthecluster,to
achieve in a predictive manner the arrival of the target. A new
cluster will be created according to the algorithm proposed
in Section once the target is detected by other nodes
outside the active cluster. Reference [] proposes to return
all member nodes of the ac tive cluster in sleep state once
they no longer detect the target aer a system time
.e
problemistoprovideawaytoproperlyadjustthesystemtime
when the target speed is variable to ensure that the cluster
nodes , by Msg
data messages, do not wake up the cluster
members that have just been put to sleep. ese messages
can be designed as the reected wake waves to the origins
clusters, which represents a considerable loss of energy by the
sleep process and repetitive activation. In Figure , node
belonging to cluster is not detecting the target at time 1
and triggers a timer
aer which this the node will return
to the sleep state, while, at time 2 and time 3,MSG
Data
messages wake up node belonging to cluster , which also
wakes node . It is obvious that the proper adjustment of
depends on the speed of the target and the density of nodes
in the network. How then can one adjust this setting in an
environment where the velocity of the target and the density
of nodes are variables?
Our algorithm solves this problem by synchronizing the
transfer to the sleep mode on listening posts MSG
data and
not on the moving target. A node cannot return to the sleep
state if it does not receive aer time
MSG
data messages,
the messages of predictions, or t he reected wake waves. is
method of synchronization will not therefore be aected by
thevariablespeedofthetargetorbythedensityofnodes.
3.6. e T ransition State Diagram
(i) Fivepossiblestateshavebeendenedinthealgorithm
asshowninFigure:sleep,CH-ready,CH,CH-for,
and member.
(ii) Initially or when there is no target activity in the
network,allnodesareinthestateof“sleep.”
(iii) When node is in “sleep” state, it receives Msg-data
packet from another node belonging to the active
cluster; node goes to the CH-For state if <
or
CH-ready state otherwise.
(iv) is the distance between the two nodes and .
(v)
is the detection range.
(vi) Only the nodes with the CH-ready state can become
CH aer their detections of the target; contrariwise,
the nodes with the CH-For state can only become
members.
(vii) A node in CH-ready state and receiving MSG-data
to another node , which is belonging to the active