International Journal of Applied Engineering Research ISSN 0973-4562 Volume 10, Number 22 (2015) pp 43029-43044

© Research India Publications. http://www.ripublication.com

Mobile sink routing and relocation strategies in wireless sensor networks

A Review

Vicky Kumar

Ph.D Scholar, Department of E &CE NIT, Hamirpur, India, [email protected]

Ashok Kumar

Associate Professor, Department of E & CE NIT, Hamirpur, India, [email protected]

Gaurav

Ph.D Scholar, Department of E & CE NIT, Hamirpur, India, [email protected]

Abstract- contribution towards WSN and is considered as a dead node

Mobility brings new revolution and challenges in the field of in the network. In such situations, other nodes around such
wireless sensor networks (WSNs). It improves lifetime of dead node have to balance the network load. The energy
sensor network by repealing network hole problem near static consumption of sensor node mainly depends on transmission,
sink. Sink trajectory has significant impact on network reception and computation of data packets. So it is very
performance in terms of network lifetime and data delay. important to optimize them.
Therefore, sink relocation methods must be wisely used to Energy efficiency is the primary goal while designing a sensor
improve performance of WSNs. Uniform energy network. As shown in the figure (1) the nodes near to sink are
consumption, low latency, load balancing and long network handling maximum packets as compared to other nodes in the
lifetime are major outcomes of mobile sink. But, designing network. So energy of these nodes will exhaust faster. As the
routing path to a mobile sink for guaranteed data delivery in dead nodes are not able to forward packets to sink, hence no
energy-efficient way becomes challenging objective. Sink further processing is possible in the network [12,13,14].
mobility requires route adjustments with respect to fresh sink Therefore network may fails, while 90% of network energy is
location which results extra energy consumption and still remaining. The idea of mobility (node or sink) has been
increased delay. Hence sink mobility insights new research mooted to addresses this bottleneck. The mobility shifts the
problems in WSNs. In this paper, we review sink relocation high traffic nodes to the low traffic area and increases their
techniques and routing protocols for handling sink mobility lifetimes. A number of techniques have been proposed on
which may help for selecting suitable techniques to get best different layer for efficient data transmission. However, this
results for different applications designing. A brief discussion survey paper mainly focuses on the sink mobility techniques
about new challenges concerned to mobile sink is also and routing algorithms for handling sink mobility.
introduced in this survey paper. The paper is organized as: In section 2, mobility requirement,
advantages, various mobility patterns and models are
Keywords: Relocation, Sink mobility, Data delivery, WSNs. discussed. In the section 3, sink mobility with fixed,
controlled and uncontrolled manner is discussed briefly. In the
section 4, dynamic sink route adjustments and data collection
Introduction algorithms are discussed. Section 5 contains the current
Wireless sensor networks (WSNs) have received much research issues and section 6 discusses the conclusion.
attention because of their peculiar properties to observe the
changes in physical quantities like temperature, pressure and
humidity. Military surveillances, seismic activity monitoring,
animal movement tracking, intrusion detection, industrial S
chemical or smoke detection, flood detection, environmental i
applications, under sea water observations, agriculture and
smart building are some applications where the sensor
n
networks are utilized with good performance [1-11]. k
Advantage of WSNs is that the tiny node used to sense the
changes in the environment or physical quantity are self
organized. They can manage their operation by using multi-
hop communication to send data. These sensor nodes co-
operate each other for forwarding data towards the sink.
Sensor nodes are portable, small and battery operated. As the
energy or battery of a node is exhausted, it ceases it‟s Fig.1. Wireless sensor network

43029
 International Journal of Applied Engineering Research ISSN 0973-4562 Volume 10, Number 22 (2015) pp 43029-43044
© Research India Publications. http://www.ripublication.com

Mobility in Wireless Sensor Networks Nomadic community mobility are used to predict new
Mobility is used for the load balancing in the network [15-21]. locations for a set of nodes in next slot [29].
In a mobile sensor network, mobility is provided by using
Springs [22], Wales [23], or by using transporters or robots
[24]. The mobility in the network may be random, controlled Sink Mobility for Network Life Time Enhancement
or predictable. The various patterns and models have been Generally sensor networks are classified into stationary and
discussed in this section. dynamic sensor networks. In the static sensor network nodes
are fixed in their respective positions and perform their
A. Mobility patterns sensing operation from the same place. On the other hand,
The patterns which help to provide the mobility in sensor nodes in dynamic sensor networks can move to another place,
networks are called as mobility pattern (e.g. mobility of whenever relocation conditions have been met. In the same
human beings, cars and fluids). Mobility pattern must be way the sink can also relocated to a new position by providing
known before designing the routing algorithms for a particular relocating conditions [30]. This section reviews different
application. So the various mobility patterns available in techniques proposed in the literature for the sink relocation
literature are discussed in this section.
A. Sink Mobility using predefined path
i. Pedestrian mobility pattern i. Joint sink mobility and routing strategy
Pedestrian type of mobility is inspired by mobility of human Luo and Hubaux [31] proposed a Joint Sink Mobility and
beings. For simulating this mobility patterns, sensors are Routing strategy (JMR) for lifetime elongation in the wireless
installed in the body of a person and the mobility is provided sensor network. Data collection is carried out by moving the
by the movement of person. The pattern has limited speed, sink regularly in circular trajectory and multi-hop
random motion (direction of movement) and protection from transmission used to save the energy of the nodes which are
obstacles. This mobility pattern is used in two-dimensional closer to sink at the initial stage of the operation. In Joint sink
only and it may or may not be able to describe the behavior of mobility model author proposed a predefined circular
a group [25]. trajectory as shown in figure (2). The sink will move with the
constant velocity in the predefined path. The node near the
ii. Vehicular mobility pattern trajectory will collect the data from their neighbors and
For simulating this mobility patterns, sensors are placed in the deliver to sink when the sink will pass through them. The
vehicles and movement is provided by the motion of the movement of sink will distribute the load among a large
vehicle. The vehicles having sensors can communicate with number of nodes and eliminate the bottleneck problem due to
each other and can detect their positions, which proved to be the fixed position of the sink. However in the proposal [31]
helpful for avoiding accidents on road. Vehicle mobility is the node near the trajectory have to buffer the data for a long
one-dimension only and it can characterize the behavior of a time and the buffering space may be exceeding the limit and
group of vehicles at highway [25]. drop data before it can be delivered to sink.

iii. Dynamic Medium Mobility Pattern Relocating

This mobility patterns inducing motion in the sensors by some
trajectory
medium like wind, water, etc. The dimensions of mobility
depend on the type of medium used. Motion may be one, two
or three dimensional [26].

B. Mobility models tr
Handlings mobility in the wireless network requires models
which can explain the characteristics of mobility patterns.
Random walk mobility model, Random waypoint mobility
model and group mobility model are three basic models which
are generally used and help to define the mobility in the Fig.2. Joint Sink mobility [31]
network [27-29]. Random walk mobility monitors the
mobility of a node up to a predefined distance or time. The
direction and speed change cannot be controlled during ii. Arbitrary moving trajectory (AMT)
motion [27]. Random waypoint model is almost same as the Saad et al. [32] suggested a plan for the sink mobility in
previously discussed random walk model. However in this arbitrary moving trajectory (AMT) in hierarchical structure. A
model pause time parameter is included during the direction cluster head is declared among the nearby nodes, which
or speed change. The fast speed and large pause time of nodes collects the data for nearby nodes. The cluster head buffers the
results in very stable network, whereas slow speed and low packets until the moving sink is reached in the range of cluster
pause time results in poor network stability [28]. head. The cluster head formations save energy by reducing the
In Group Mobility Model movement of nodes is correlated number of transmissions for data delivery towards sink.
with respect to each-other and new location by node is found Energy balance among the nodes is an open research issue in
for next slot. This type of model is known as group mobility such type of algorithms. Duty cycling becomes essential to
models. Exponential Correlated Random mobility model and avoid the energy hole problem in such cases.

43030
 International Journal of Applied Engineering Research ISSN 0973-4562 Volume 10, Number 22 (2015) pp 43029-43044
© Research India Publications. http://www.ripublication.com

iii. Multiple sink placement for event delivery are the important characteristics of a WSN.
Marta and Cardei [33] reduces the deficiency of joint mobility There are number of applications where the event-driven
sink [31] by increasing the sink numbers and provide them approach is more convenient than time-driven approches (e.g.
separate trajectory to collect the data as shown in figure (3). Intrusion detection, seismic activity monitoring, animal
This results in reduction of waiting time for data delivery and movement tracking). In the event-driven operation a node
improvement in latency. In this model, sink placement is transmits data only if it find any event in the network. Then
important for the load balancing of the network. The arrows in the node transmits the data to the nearby sink through
the figure (3) show the direction of motion of sink in the multihop transmission. Author proposed a trajectory for the
predefined trajectory. The nodes passes data to movable sink sink mobility which reduce the delay time for the data
when it passes through them. Each mobile sink is responsible delivery and increase the network lifetime. Author proposed
for the data gathering in a defined region. The mobile sinks network center as the best position for the sink placement.
gathers the data in short time period and improved the quality This position results in the shortest path data transmission and
of services (QOS). proper load balancing in the network.

B. Sink mobility by using Energy awareness

These algorithms consider the energy of nodes for the sink
relocation. The one hop neighboring nodes to sink played
important role in energy aware sink relocation techniques.
Various techniques for sink relocation using energy factor are
discussed below.

i. Moving destinations (Move Dest) Scheme

Limin Sun. And Yanzhong Bi [37] partition the area near the
sink in eight fan shaped sectors. The node (a node) having
maximum energy in a defined area (eight fan shaped) is
named as Move Dest (figure 4). The part of eight fan shaped
Fig.3. Multiple mobile sink [33] containing the Move Dest is called as Dest sector. A threshold
value is set for minimum energy that is required for nodes to
iv. BIORM operate in a network. Any node in the network having low
Sumit Kataria and Aarti Jain [34] proposed Bio inspired energy than this threshold value is called as quasi hotspot. The
optimal relocation methods (BIORM) for sink. This model sector where any hotspot found is declared as miry sector. The
uses cluster head (CH) to collect the data from their sector having no hotspot is called as clean sector. The
neighboring node. The CH is a special node having capability movement of sink is carried along the straight line from the
to move in its vicinity. The CH can be considered as a sink. current position of sink to Move Dest. The next position of the
So the number of sink present in the network is equal to the sink is the intersection point of transmission radius of sink and
number of CH present. Each sink is responsible for own straight line. Relocation of sink near the Move Dest starts
cluster operation only. Relocation of sink is performed by Bio utilizing the energy of Move Dest and it avoids the bottleneck
inspired Digital Hormone Model. Sink generates or forward problem near the sink.
hormone to nodes that present in its vicinity or
communication range. For relocation, sink will collect the
information about local residual energy, global behavior and
maximum traffic from these hormonal messages. The
massages help sink for taking future decisions with regards to
it‟s relocation.

v. Adaptive motion control (AMC)

A. Somasundara [35], proposed an adaptive motion control
(AMC) for the sink mobility. The motion of sink is correlated
with traffic in the region (i.e. how much time sink will spend
in a defined region). The heavy traffic regions are defined
earlier and the sink is advised to stay there for longer time as
compared to the region of low traffic. Hence the speed of sink
is slow at high congestion region. Slow speed gives large time
to collect the packets at the highly congested area. However,
high speed at low congestion region reduces the unnecessary
time delay.
Fig.4. Sink Relocation [37]
vi. Sink mobility in the event-driven sensor network
Bin Wang [36], proposed sink mobility in the event-driven Six cases may be possible due to the energy levels of nodes
sensor network. Enhanced network lifetime and reduced delay nearer to Move Dest [13]. Number of sectors shown is eight,

43031
 International Journal of Applied Engineering Research ISSN 0973-4562 Volume 10, Number 22 (2015) pp 43029-43044
© Research India Publications. http://www.ripublication.com

which is not limited to eight only. The partitioning of sectors starts rotating in a circular path with communication radius.
can be further increased with the increase in node density. Six moving positions are shown in figure (5b). At each
position sink checks the traffic and finalizes the relocation
ii. One step relocation (OSR) position by getting the highest traffic among all rotating
Yanzhong Bi [38] proposed two sovereign methods, one step points. The relocation process is repeated, when residual
and multistep moving to relocate the sink (figure 5a). The idea energy becomes less than threshold value.
behind these methods is to monitor the heavy traffic nodes in
the network. Moving decisions can be made without the
knowledge of node energy distributions and network
topology. In the sensor networks, data collection is carried out
by multi-hop fashion. Therefore, sink has the information
about the residual energy of each node. The highest residual
energy node is considered as moving destination and the sink
is placed in the communication range of moving destination.
In the one step moving method the sink is moved in the
communication range of the Moving destination in one step
regardless the distance between them. As the moving
destination (highest residual energy node) is present in one
hop distance to the sink, hence it spends large energy
compared to other nodes by forwarding the packets of other sink
nodes. When the energy of moving destination becomes less
than the predefined threshold energy, a new moving
destination is created.

Point visited Moving Trace

Data flow Common nodes

Fig.5b. Multi step relocation [38]

iv. Energy aware sink relocation (EASR)

Moving Chu-Fu Wang and Jau-Der Shih [39] proposed a practical and
Destination mathematical model to enhance the network lifetime using
Comm. Energy-Aware sink Relocation (EASR) method. In proposed
Range protocol transmission distance of the sensor node is bounded
with their residual energy. Nodes having small residual energy
transmits for short distances only and save energy for future
processing As shown in figure (7), the nodes having battery
level B transmit up to distance „r‟ and transmission range
Fig.5a. One step relocation decreases with decrease in their battery level. Sink relocation
conditions are proposed by Chu-Fu et.al [39]. Hence
relocation process is carried out, only if relocation conditions
iii. Multi step relocation (MSR) are met. Maximum capacity protocol (MCP) is used to find
Multi step moving scheme takes numbers of steps or iteration the healthy path for data transmission [40]. If any node in the
before finalizing the next position of moving destination path has the battery level less than B/2 or the average residual
(figure 5b). One step moving schemes is suitable in scenario energy of a set of nodes N which are the part of the path is
where the speed of sink is not limited. However if the speed of less than B/2, sink relocation is processed.
sink is limited, then multi step moving scheme is used. Sink Moving destinations need to be calculated a priori. Author
can be moved during data collection and it moves only within proposed 4 positions for the moving destination , , ,
it‟s communication range [38]. So we need intermediate and as shown in figure (6). Suppose each moving
moving destination path to reach the final moving destination, destination has neighbor subset (figure 6). Assuming a
which can be anywhere in the network. Sink selects the weight value is associated with each neighbor subset (1
highest energy node as an intermediate node within it‟s I 4 ), = min {c |u }, where c represents
communication range in the direction of final destination and
moves towards intermediate destination. Afterword sink again the maximum capacity value of path . The position having
finds new intermediate moving destination for relocation. This highest weight value is fixed for relocation. This method
process is repeated until sink reaches at the final moving provides a better relocation among four possibilities for
destination. When sink reaches at the moving destination, it relocation and avoid bottleneck problem of energy depletion

43032
 International Journal of Applied Engineering Research ISSN 0973-4562 Volume 10, Number 22 (2015) pp 43029-43044
© Research India Publications. http://www.ripublication.com

near the sink. This method also provides healthy path for the Region (quadrant) having lowest energy nodes for the one
data dissemination. period is called quasi-hotspots in the eight half-quadrants.
Depending on the position of such nodes sink decides the path
or trajectory to move. Generally two cases are possible, the
moving destination may be close to sink or it may be far away
from sink.When the Move Dest is far away from sink, the
trajectory will try to avoid the quasi-hotspot. If it is not
possible to avoid the quasi-hotspot from the path, the
minimum-influence position selection algorithm (MIPS) is
followed [41].
In other case, when the Move Dest (moving destination) is
near to the sink, a square hopping mechanism is used for sink
relocation. Square hopping mechanism requires four points
selection on a circle around the Move Dest, whose radius must
be less than communication distance of the sink (1-hop).This
is carried out to move the sink one hop away from the node
having highest residual energy. Henec this highest residual
energy node can gather the data of other nodes. Nodes having
energy less than a certain threshold value is considered as
dangerous nodes and added to the blacklist. These blacklisted
nodes are not allowed to forward any data. Otherwise, these
Fig.6. Sink relocation process in EASR [39] nodes can create network hole. Sink uses two threshold values
and to decide whether a node is blacklisted or not,
where represents the data gathering period for the same
MoveDest and represents the number of data messages
received from the same MoveDest. For deciding a blacklisted
node the value of must exceed a predefined threshold
value and the value of must be reduceded from
expectation for same Move Dest.

vi. Integer linear program (ILP)

Integer linear program (ILP) is proposed by Gandham et al.
[42] to find the new location of the sinks. Flow-based routing
Transmission range „r‟ Transmission range „r/2‟ protocol (FBR) is used for data delivery. ILP is used to
minimize the energy consumption in each round by a sensor
Fig.7. Adaptive transmission range selection node. ILP solutions are helpful to find out the new location of
sink. At the start of each round, the sink is placed in a new
position to collect data for the round. The sensor network is
v. Half -quadrant based moving strategy (HUMS) considered as a graph G (V, E), where V = U and is
Half-quadrant based moving strategy (HUMS) [41] is the set of sensor nodes and is the set of feasible sites. E
proposed to move sink autonomously. Mobilty of sink is
towards the higher residual energy nodes. So these nodes can V V is the set of wireless links [41]. In integer linear
become data forwarder for nodes having low energy. HUMS program (ILP) a variablre corresdponds to integer 0-1 for
is incorporated with the data-gathering protocols. Data- each l Such that:
gathering protocols are used to collect the information for the
node having highest and lowest residual energy at their = (1)
present locations. The position of moving destination (Move ILP includes the equations for number of packets a particular
Dest) is changed after every data-gathering period. The node (i) transmits to a neighboring node (j), energy for the
position of new moving destination depends on the location of reception and transmission for packets, maximum number of
nodes having highest and lowest residual energy. The motion feasible sites and guaranteed delivery of packets. ILP provide
of sink is constant and only one communication distance is better results if the sink is placed at feasible site. ILP provides
travelled in one round towards the moving destination. If a the complete information about the parameters which
new location of Move Dest is found before the sink reached at enhances the network lifetime.
Move Dest, the sink alters the path towards the new location
of Move Dest (sensor node with highest residual energy in the
network). Move Dest changing is possible, if there is Dynamic sink route adjustment & data collection
minimum one unit energy difference. algorithms
Moving decision is held by the sink and is placed at the origin. Sink mobility regularly changes the location for data
Eight half-quadrants are considered around the origin of sink. gathering. Therefore, dynamic route change are desired for the

43033
 International Journal of Applied Engineering Research ISSN 0973-4562 Volume 10, Number 22 (2015) pp 43029-43044
© Research India Publications. http://www.ripublication.com

efficient data gathering. Large geographic sensing area and for the location updates of mobile sinks in more energy
limitted energy of nodes makes data collection challenging efficient way. GA provides quick reporting to sink in the
[43-47]. Existing research shows, mobile sink should pause emergency. Only GA is responsible for the mobile sinks
for short time in low traffic area and pause for long time in tracking. Unlike TTDD, the grid construction is one time
high traffic area for best results [48-51]. Mobile elements are process in MSDD. Whereas, in case of TTDD each query
also proposed for data collection which have better results needs a new grid structure resulting in high overhead. MSDD
than multihop routing techniques [52-53]. provides reduced data reporting time and low energy for data
Dynamic route adjustments protocols have better performance delivery.
compared to fixed routes. However, dynamic route adjustment
requires extra energy. Dynamic route adjustment algorithms iii. Grid-based Energy
have been discussed in this section. Efficient Routing (GBEER) [56] is a modification of TTDD
protocol [54]. GBEER forms a single grid source for all
sources. The location of each node must be known a priori to
A. Hirarachical Approaches form a grid struture. Sink transmits the data requests, while
Major problem in the wireless sensor network (WSN) is high the data announcement are generated from the source. Data
traffic near the sink which is terned as hot spot area. Sink announcements always propogate horizotally and data
mobility can be used to reduces the problem. However, in requests propagate vertically along the grid, assure these
order to increase the performance, the hierarchical routing packets met at a crossing point. Now the position of sink can
approach are mostly used. These approaches help to distribute be forwarded to souce node and direct data transmission is
the load among a group of nodes.Various hierarchical possible. GBEER reduces the high overhead of TTDD. Nodes
approaches used for sink mobility have been discussed below. forming grid struture depletes energy faster compared to other
nodes, hence grid nodes replacement is necessary. Hexagonal
struture can also be used for data gathering. Hexagonal Path
a. Virtual-grid based approaches Data Dissemination (HPDD) and Honeycomb Tessellation
In this approach, network is partitioned into some structures (HexDD) are the examples for hexagonal grid struture [57].
such as grids to handle the sink mobility. The network
structure helps to distribute the load among a group of nodes. iv. Virtual Grid-based Dynamic Routes Adjustment
The shape of structure may be rectangular, circular or (VGDRA)
hexagonal. Virtual grid approaches are widely used in the Virtual Grid-based Dynamic Routes Adjustment (VGDRA)
target detection applications. Various virtual grid based [58] scheme collects data periodically. The algorithm provides
approaches are discussed below. balance between the energy consumption and data delivery.
VGDRA does not use multiple sinks or super nodes to collect
i. Two-Tier Data Dissemination (TTDD) the data. Proposed technique maintains the optimal route for
Two-Tier Data Dissemination (TTDD) [54] delivers data for the mobile sink from each sensor node. In VGDRA, sensing
multiple mobile sinks. The position of nodes is fixed and field is divided into virtual grid of K-cells having equal size.
known, while the sinks are mobile. Data for event is always A virtual backbone network is constructed for all cell-header.
generated by a source node. The source node is considered as Nodes closer to centre are considered as cell-header and
starting crossing ponit for grid and grid overlay network is collect data for their member nodes. The delivery of data to
formed. Grid points acts as disseminating points. A mobile mobile sink is through the virtual backbone network. VGDRA
sink transmits the query and selects a dissemination point near reduces the cost for dynamic re-route adjustments. The energy
the mobile sink. This query message is routed through the cost and end to end delay in data delivery phase are also
overlayed network toward the source (specified in message). improved in VGDRA. Only a few cell-headers participate in
The query packets contains the information about the static re-route adjustment, so system performance is improved in
node near the mobile sink termed as primary agent. When sink terms of route maintenance cost.
moves out of the range of primary agent, sink chooses an
immidiate agent for the robust delivery. Source reply for the b. Cluster-Based Approaches
query through overlayed network and pass data to the neartest Cluster formation is the most efficient approach for enhancing
disseminating point and finally to the primary agent by network lifetime in the wireless sensor networks. A cluster
disseminating node. As the sink changes it‟s position, new head (CH) is selected in each cluster. All decisions are made
primary agent will be selected and old one is disposed. In by CH which responsible for data gathering in a cluster
event sink move away from the old one, new dissemination region. The different cluster-based approaches used for sink
node can be found by repeating the same process. TTDD mobility are discussed below.
provides high overhead due to the grid formation for each
query. This drawback is addresed by the MSDD [55]. i. Mobile sink-based routing protocol (MSRP)
Mobile Sink-based Routing Protocol (MSRP) [59] presents a
ii. Multiple mobile sinks data dissemination (MSDD) data collection technique in which the cluster head (CH)
Multiple mobile sinks data dissemination (MSDD) [55] collects data for it‟s members and wait for the mobile sink
modifies the TTDD and grid based mechanism for the until it reach in CH‟s vicinity. Data is handed over to the
multiple mobile sinks updates. MSDD uses global agent (GA) mobile sink as it comes to the vicinity of CH. The cluster
along with the primary agent (PA) and immediate agent (IA) heads uses TDMA techniques for fixing their slots for data

43034
 International Journal of Applied Engineering Research ISSN 0973-4562 Volume 10, Number 22 (2015) pp 43029-43044
© Research India Publications. http://www.ripublication.com

transmission. Sink mobility is controlled and residual energy registration cost, but at the expense of high delay to reach
of neighboring CH‟s is known. However, the possibility to packets at sink.
attend each CH by sink cannot be guaranteed. Hence MSRP is
not beneficial for the real time applications. However, it is C. Area-Based Approaches
very effective in energy-efficient applications. Routing approches which provides a specifed area for the data
collection in a network are called as Area-based approches.
ii. Energy-Efficient Mobile Sink Routing Hot spot problem in the wireless sensor network can be
Energy-Efficient Mobile Sink Routing Algorithm [60] is reduced by increasing the size of data collecting area.Various
based on LEACH protocol [61]. Cluster formation is carried algorithms in this catageory are discussed below.
out by LEACH and addition of enforced TDMA schedule for
nodes to send their data in EEMSRA. The mobile sink i. Line-based data dissemination (LBDD)
broadcasts the visiting schedule to CH a priori and updates the Scalibility and hot spot are two major problems in the wireless
routes to reach the cluster heads. Sink select the highest sensor networks (WSN). Line-Based Data Dissemination
energy CH as next visiting point for data collection in next (LBDD) [66] protocol tried to address these issues. A vertical
slot. The known trajectory and controlled mobility of sink line of width „w‟ is used to divide the sensor network into two
consumes low energy. Since EEMSRA works on the MAC parts (figure 9).Vertical line is placed in the center so that
layer, so cannot be used in wide areas. However, hot spot each node can approch the line easily. Line of width „w‟
problem is diminished by changing CH with member nodes contains some nodes which are called as in-line nodes and rest
regularly. nodes are called as ordinary nodes. The line acts as rendzvous
area for the data storage. Each source node in network
iii. Backbone-based Virtual Infrastructure (BVI) forward their data towards line only. The qurey is generated
Backbone-based Virtual Infrastructure (BVI) [62] scheme by the sink and transmitted towards the line and line replied
proposed a single-level multi-hop clustering for the data with data that is requesd in the query. Source nodes can easily
dissemination. BVI reduces the number of clusters and results access the line and it helps to reduce the overhead during the
low network overhead. HEED protocol [63] is used by the data gathering. LBDD uses broadcasting to forward the
BVI scheme for the clustering. Nodes having higher residual quries. The in-line width should be enough to handle the hot
energy are considered as cluster heads (CHs). One of the CH spot problem. A simillar approach called as railroad [67] is
is appointed as root CH and assumed that it is responsible to used to control the flooding for query dissemination in the
handle the sink mobility. When the mobile sink enters the network. Railroad uses a unicast transmission of query on the
sensing area, it registers itself with the closest CH through an rail. As the query get the requested data, direct transmission of
agent node (as the case in the cellular networks). The host CH data towards the sink.
updates the location of sink regularly to root CH and rest of
CH‟s. Roaming of sink within the cluster is not intimated to ii. Path constrained model ( PCM)
root CH, this mobility is handled by the CH of the cluster. If Shuai Gao and Hongke Zhang [68] presented a data collection
the sink leaves the cluster, a new registration process started technique for a mobile sink in a path constrained model. The
by another agent node. Root CH becomes the focal point for path and velocity for mobile sink is fixed as shown in figure
the route adjustments and exhausts it‟s energy quickly. (10). The sink is placed on a vehicle and move in predefined
trajectory with constant speed periodically. As the path ends,
iv. Hierarchical cluster-based data dissemination sink reverse the direction and move in same path periodically.
(HCDD) In the presented model number of sub sink and one mobile
Hierarchical Cluster-based Data Dissemination (HCDD) [64] sink is used for data collection purpose. Monitoring region is
provides a routing agent to mobile sink. This routing agent divided into parts. The region which is in direct range of
handles updates of movable sink. Routing agent broadcast, the mobile sink (M) is considered as direct communication area
new location updates to it‟s nearby routing agents. High sink (DCA) and the region which is not in direct communication
mobility results in more energy consumption in HCDD. Data range of mobile sink is called as multi-hop communication
delivery paths are not optimal due to restricted updating area area (MCA). Sensors present in MCA are known as members
for sink location, which results in high latency. and relay their data to sub sink for the final delivery to mobile
sink (M). The sensor nodes which are present in the DCA can
v. Cluster based mobile sink exploration (CMSE) directly send the data to mobile sink, when it comes in the
Wei-Cheng Chu and Kuo-Feng Ssu proposed a cluster based range of these sensor nodes. Since path and speed of mobile
mobile sink exploration (CMSE) [65] routing technique to sink remains same, therefore time taken by each sub sink to
find the location of mobile sink. CMSE divides network into deliver their data to mobile sink remains constant. Throughput
clusters and CH is assigned to each separately. Among these of the model depends on upper limit of sub sink for data
clusters, some are used to find the location of sink. Such gathering in one round and the number of sensor nodes
selected clusters are called registration web. When transmitted sending their data to a sub sink.
packets reach to any cluster which is a part of web, they are
guided towards sink. Web size decides registration cost and iii. Ring Routing
ease of sink exploration. Larger web size results in high Ring routing [69] uses very small area (one node width)
registration cost, but packets are routed towards sink with (figure 9) in network for the location updates of mobile sink.
minimum delay. Whereas, small web size results in low Involvement of limited nodes (ring nodes) in sink location

43035
 International Journal of Applied Engineering Research ISSN 0973-4562 Volume 10, Number 22 (2015) pp 43029-43044
© Research India Publications. http://www.ripublication.com

updates, results to low energy consumption. Ring nodes forms region where connectivity is not possible. Data mules motion
closed loop and enclose center of network which is predefined is considered to be uncontrolled and unpredictible. When the
globally. Mobile sink is advised to send the location updates data mule comes into contact to any node in the network, it
towards network center and these updates are captured by ring can get the data for delivery. The density of data mules
nodes. Source node collects the sink location by requesting depends on the area coverd and number of nodes used for
ring nodes. As source gets the fresh sink position, it can sensing. The data delivery delay and buffer size of data mules
dirctly transmit data using geograhic techniques. Ring routing are the challenges for the future research. The buffering
provides an energy-efficient mechanism for sink location capability of data mules is limited. A full buffer my results in
updates. But, initially ring construction is time consuming packet dropping before delivery them to sink. Data mules
process. provide load balancing and contribute to increase the life time
of the network [71].

v. Multiple enhanced specified-deployed sub-sinks

(MESS)
Multiple Enhanced Specified-deployed Sub-sinks (MESS)
[72] scheme form a virtual strip in the centre of sensing field.
This virtual strip contains advanced nodes (sub-sinks) having
more memory to store data at equal distances. These sub-sinks
with proper paths acts as rendezvous points for the mobile
sink. The data is stored by the sub-sinks. The query is flooded
in region where sub-sinks (rendezvous points) are placed. On
Fig.8. Mobile sink routing techniques receiving the query from mobile sink, sub-sinks forwards the
data by using the geographical forwarding scheme.

vi. Virtual Line-based Data Dissemination protocol

(VLDD)
Virtual Line-based Data Dissemination protocol (VLDD) [73]
is proposed to overcome the deficiency of Geocasting and M
Geocasting. The VLDD proves to be more energy efficient
and have better data delivery ratio than M Geocasting. Author
proposed a Virtual Line Structure (VLS) to store the data of
source, unlike flooding of data to a region as the case in [74].
The sources save their data in VLS and the sink retrieves data
from VLS when required (figure 11). As the number of sinks
increases in VLDD, data delivery ratio decreases
Fig.9. LBDD (left) and Ring Routing (right)

Sink trajectory MCA

M DC
A

MCA

Figure 10 Data collection by Mobile sink in Path

Constrain Model (PCM) [68]
Fig.11. Virtual Line-based Data Dissemination protocol
(VLDD) [73]
iv. Data Mule
Large and sparsly deployed sensor networks suffer from .D. Tree Based Approaches
connectivity problem. To overcomes this difficulty Data Mule Approaches which disseminate the mobile sink location in a
[70,71] architecture is presented. Data mules are the mobile tree structure (from root to leaves) are called as tree-based
nodes and these data mules are used to gather data from the approaches. The root node disseminate update for mobile sink

43036
 International Journal of Applied Engineering Research ISSN 0973-4562 Volume 10, Number 22 (2015) pp 43029-43044
© Research India Publications. http://www.ripublication.com

location to it‟s leaf nodes. Techniques which based on tree i. Direct Diffusion (DD)
approach are discussed below. Direct Diffusion (DD) [78], provides sink location updates to
the sensor nodes in the network. Sink regulary sends updates
i. The scalable energy efficient asynchronous by flooding messages throughout the network. Each node in
dissemination (SEAD) the network form a shortest path towards the sink for data
The scalable energy efficient asynchronous dissemination transsion in more energy efficient way. Sink is always
(SEAD) [75] technique is helpful in data delivery where respossible for route formation and data transmission. Sink
multiple mobile sinks are used. A static node near the sink is transmits the query and interested nodes respond in the
used for communication to or from source node to mobile reverse path to the query. Direct diffusion protocol is event-
sink. The static node/Access node is responsible for sink driven case. Flooding for sink updation increases the traffic
monitoring. When the sink is out of range of the access node, and collission in the network, which increases the energy cost
a new access node is formed. A multi-cast dissemination tree through retransmission.The performace of data gathering by
is formed to send data from source node to various access Direct Diffusion protocol is affected by the sink position, area
node present in the network. Hence data delivery is of network, number of sensor nodes and communication
guaranteed using SEAD protocol for mobile sinks in energy network topology. DD protocol is not used in the case where
efficient way. time boundation (real time applications) is must for data
delivery. But the data delivery can be expected because there
ii. Quad-tree Based Data Dissemination (QDD) are number of paths that can complete the data delivery even
Protocol in the worse case.
Quad-tree Based Data Dissemination (QDD) Protocol [76]
divides the network into consecutive quadrants The centre ii. Rumor Routing (RR)
point in each quadrant becomes second-tier node. Each Rumor Routing [79] is presented for case where the number
quadrant is further divided into smaller quadrants until the of events are very rare. Flooding in such cases is not helpful.
resolution of second-tier nodes become sufficient for the fast In rumor routing the messages are flooded in the event
access to virtual structure. The centre points of quadrants happening region only. This type of routing needs geographic
gather data announcements and queries until they become locations clearly.
rendezvous. The overhead in QDD is very less compared to
proposed algorithms in the paper, but the solution for hot spot iii. Adaptive flooding-based location update protocol
in QDD is not proposed yet. (ALURP)
Wang et al.[80] proposed a methods to handle the sink
iii. Virtual Circle Combined Straight Routing (VCCSR) mobility. Nodes in the network transfer their data to sink. If
Virtual Circle Combined Straight Routing [77] is a converge- sink location is not known to nodes, delivery of data is not
cast tree algorithm that form virtual structure by using a possible. When the sink moves to new location, each node
virtual circle and straight line. A set of nodes near to the must know the new location of sink as given by the Flooding-
virtual circle and straight line are declared as cluster heads. A based Location Update Protocol (FLUP) [81]. To update
backbone network is formed by these cluster-heads. The sink every node about the sink location is highly energy consuming
moves around sensing field and maintain communication with process. Wang el al. proposed Location Update-based Routing
the cluster-heads present at border for data collection. Protocol (LURP) and Adaptive Location Update-based
Although VCCSR reduces the re-route adjustment cost, but methods for the sink location updation. Routing protocol
CH at center becomes the focal pont in re-route adjustment (ALURP) is improvement of LURP [80]. Author suggested
and depletes energy quickly. that it is not neceessary to update the position of sink to each
and every node in the network. Only a few nodes nearer to the
B. Non -Hierarchal Approaches sink are sufficient for data delivery. Therefore, only a
The nodes have same capability and functionality in non- particular region (Virtual Centre radius) around the sink is
hierarchical approches. No special structure (e.g. virtual grid, choosen for the sink location updates. Initiation of
cluster) required for the operation. These approches save time transmission by a node requires two steps. Firstly, the node
and energy for the structure formation. Various routing have to send data to node which is in the radius of virtual
algorithms for non hierarchal approaches are discussed below. centre (VC). Secondly, it needs to find the dissemination node
(having sink location updates), which pass the data to sink
a. Flooding Based Approaches through any topology-based routing protocol. The radius of
Broadcasting nature of wireleess transmission is used for sink VC have impact on the network life time. As radius of VC is
loaction updates or data delivery to sink. Since sink is considered to enough large, the number of nodes for updation
movable, so data is transmitted all over the network for are increased and consume more energy. If the VC radius
confirmed data delivery to sink. Flooding is the simplest radius is considered to be small, new VC selection rate
routing techniques, but having highest energy consumption increases due to movement of sink. Henec radius of VC must
among all types of algorithms. Various flooding-based be choosen carefully by considering the overall network
approches are discussed below. depoyment. ALURP have better results than LURP and FLUP
in term of network life time.

43037
 International Journal of Applied Engineering Research ISSN 0973-4562 Volume 10, Number 22 (2015) pp 43029-43044
© Research India Publications. http://www.ripublication.com

iv. Gradient based broadcast (GRAB) the data. These rendezvous points are able to gather the data
Gradient Based Broadcast (GRAB) [82] uses credits value and of neighboring nodes by using multi-hop communication. The
decrease credits regularly with each packet transmission rendezvous points are basically data collector with sufficient
towards sink. Nodes having less credits value than their amount of memory and buffer the data until sink comes in it‟s
neighbors are eligible for the transmission (figure 12). Initially range. When the sink comes in range of these rendezvous
large numbers of paths are formed by broadcasting. Therefore, points they deliver data to the sink. After data delivery buffer
Intermediate nodes forward the packets to large number of is empty and ready for data collection for the next round.
eligible nodes and consume maximum credits. Each packet Nodes which send their data to rendezvous points are aware
contains three fields: (i). Credits at the originating node. about the sink tour. Nodes are also aware about their distance
(ii) cost to sink at original node. (iii) „U‟ Budget from the rendezvous point. Hence data forwarding is adjusted
consumed before the current node. First two fields are same in such a way that data can be delivered to rendezvous point
but third field may change and depends on transmission cost. on time. Presented algorithms reduces long multi-hop path for
Initial credit is the tunable parameter and credit value can many nodes in the network. Also, the rendezvous points are
always fixed, so the topology adjustments are not required. RP
be increased for better results.
helps to save the energy and increases the possibility of
successful data delivery. Controlled sink mobility and in-
v. Mobile Geocasting (M Geocasting)
network data gathering collectively provides a good load
M Geocasting technique [83] is recently proposed for mobile
balancing, energy saving and low delay.
sink data gathering. Mobile sinks locate themselves in a
defined region. Each source floods their data towards defined
iii. The Geographic Hash Table (GHT)
region without any routing path like the geocasting [84]. To
The Geographic Hash Table (GHT) [88] uses data centric and
handle the sink mobility outside the defined region foot-print
geographic routing together. GHT uses hash functions for the
chaining scheme is used. M Geocasting can be used only for
unique location of the sensing area. Nodes deliver their data
small area networks. As area increases, node density also
close to these locations and when sink needs the data, it sends
increases and results in flooding storm [85]. Flooding storm
query to these locations to retrieves the data. GHT reduces
increases energy consumption rate and low data delivery
overhead in data gathering. GHT suffers with hot spot
problem and node changing on hot spot is not possible until
b. Geometric-Based Approaches
the hash function is changed.
Geometric approaches eliminate undesired flooding of
messages in unwanted region. Query transmission to
iv. Double Cross (DC)
particular area may be possible, If the location of event is
Sink mobility in large sensor networks presents new
known.. Hence geometric approaches play important role in
challenges for the data delivery. A Double Blind Data
the wireless sensor network, where energy is bottleneck for
Discovery using Double Cross is presented for the solution of
whole network. Various geometric approaches which are used
above mentioned problem in [89]. A Random Line Walk
in the sink mobility are discussed below.
Mechanism (RLW) is used for message forwarding and
Double Cross (DC) mechanism is proposed for further
i. Geographical Cellular-like Architecture (GCA)
efficient data delivery. Double Cross mechanism is based on
Geographical Cellular-like Architecture (GCA) [86] is a
geometric property of a planar where the probability of
cellular like hexagonal virtual structure to handle the mobile
intersect in a planar is 99%.This technique doesn‟t require
sink. A node at the centre acts as header node (like agent node
node‟s geographic location or informational direction. But the
in cellular) and rest nodes in the hexagonal are treated as
nodes must know the distance of neighboring nodes within
members. As the mobile sink enters in a new hexagon, sink
their communication range.
will inform old header about new header and which helps to
re-route the packets. GCA stops the sink location flooding
C. Overhearing-Based Approaches
massages, but overhead increased because of long re-route of
The packets or messages are regularly transmitted in the
packets.
network using multi-hop fashion. The wireless
communication has the feature of overhearing the packets of
ii. Rendezvous Points (RP)
neighbors. So nodes have chance to find the route of these
If sink itself collects data from every node in the network,
packets. Hence node can overhear and learns the routes
pausing time for the sink at heavy traffic becomes large which
without any energy expenditure. This technique is used for
increases the latency for the sensor network. Increased latency
finding the location of mobile sink. Various overhearing based
is very harmful for real world application, where delayed data
techniques are
have no use. One round for data collection takes large time
and performance of system degrades.
i. Data-Driven Routing Protocol (DDRP)
Guoliang Xing and Minming Li [87] proposed a rendezvous
An Efficient Data-Driven Routing Protocol [90] is presented
algorithm to gather data at RP points in the network and
for the learning routes from the broadcasting nature of the
where sink has to visit only these points for data collection.
sensing nodes. DDRP reduces the overhead for data collection
Current research focuses on mobility and short range
in sink mobility sensor network. In DDRP each packet records
communication for the energy efficient sensor networks.
the distance between sender and receiver. Data packets
Author [87] proposed some data collection points (rendezvous
passing from neighboring nodes help to find the routes to the
points) in the network where the mobile sink is able to gather

43038
 International Journal of Applied Engineering Research ISSN 0973-4562 Volume 10, Number 22 (2015) pp 43029-43044
© Research India Publications. http://www.ripublication.com

sink by overhearing. If any sensor node is not able to know limited applications. Researcher may focus to reduce the cost
the route, it will transmit the data using random walk for providing mobility in WSNs.
mechanism until the packet reaches to node having the Sink Relocation: To decide conditions and new positions for
information about the sink route. DDRP detect the routes at sink relocation is a challenging task. After sink relocation the
some extra energy expenditure. packet transmission path is totally changed and hence distant
place relocation results in the heavy data loss. Some energy
ii. Elastic Routing aware methods for sink relocation have been proposed in the
Elastic routing [91] is a geographic routing with the literature, but recovery of packets during the sink mobility is
combination of overhearing nature of wireless still a challenge.
communication, used in the moveable sink sensor network to Data Storage Limits: Data collection by mobile sink is a time
discover the routes for mobile sink. Initially source finds the consuming process. Nodes must have enough data buffering
location of sink through some location services and forwards capacity, because sink may take considerable time to reach the
the data through greedy forwarding [92]. Nodes which node. Larger buffering capacity node can hold data for long
forward the data, neighboring nodes of these forwarding time and avoids data loss. The data storage may be at cluster
nodes overhear the packets and find the route of sink by heads or rendezvous points [87]. Researchers may focus to
overhearing the traffic. As the sink changes the position, last increase data storage capacity which can compensate sink
node of the route gets the new position of mobile sink. One speed to cover the network area.
hop nodes of this node will learn the sink location by Multiple Mobile Sink: For large density network the idea of
overhearing. Hence each node gets updates for mobile sink by number of mobile sink [54-56] is introduced for data
one hop overhearing. gathering in a quicker way. Increased mobile sinks may
confuse the nodes (which are in range of more than one sink)
iii. Geographic and Energy Aware Routing (GEAR) for choosing next node for data transmission. Routing
Geographic and Energy Aware Routing (GEAR) [93] is a protocols proposed for the single mobile sink are not suitable
geographic location and energy aware technique to for such cases. Therefore, design of new routing protocols for
disseminate the query in specified area. Nodes having larger the multiple mobile sinks is required.
residual energy and closer to the sink are selected for the data Mobile sink Trajectory: Various algorithms have been
delivery to the mobile sink. Nodes in GEAR learn the cost proposed to fix the trajectory of sink [31-33]. The trajectory
(number of hops) for data delivery to sink by overhearing of and the speed of sink managed centrally. But changes in the
neighbors. There are two phases in GEAR algorithm. Initially, network field are not managed centrally, hence some
packets are forwards towards the sink region and in the distributed solution need to be provide to tackle these dynamic
second step when the packet reach in the sink region, the changes in the network. Change in the topologies may be
packets are transmitted by restricted flooding (small area made adaptive to upgrade the system performance.
network) or by using recursive geographic forwarding (dense Speed of Mobile sink: Design of routing protocol is directly
network). linked to the speed of mobile sink. To store data at different
parts of the network and reaching time of mobile sink on those
points compensates the data loss due to limited storage
Open Issues and Research Challenges capability of the sensor nodes. Researcher may focus on the
Sink mobility brings a new revolution in the field of wireless speed of sink because routing protocols require updates for
sensor networks. Sink mobility completely solves the mobile sink location and time constrain for data collection.
bottleneck energy issues near the static sink. But handling of Security: Application of wireless sensor network in military
various parameters after the sink mobility is still a challenge application (on border) requires security. Multiple mobile sink
for the current research. A brief discussion on these makes number of traffic areas. So a hacking node may take
challenging issues is covered in this section. advantages and start asking to nodes for data. Since the traffic
Mobility: Providing mobility to sink is a challenging task. As of packets in different directions may fools the nodes. The
mobility is energy consuming process and use of robots or researchers may focus to design such security algorithms
vehicles increases cost. Geographic conditions always affect which can handle the multiple mobile sink in the network.
mobility in WSNs, hence mobility can be provided into

Table 1. Mobile sink strategy comparison

Scheme Number of Reactive/ Sink movement Data Aggre- Sensor Network Sink
mobile sink Proactive trajectory gation Location Structure speed
JMR [31] One Proactive Circular No Known Flat Constant
AMT[32] One Proactive Planned No Known Hierarchical Constant
MSP [33] Multiple Proactive Planned Yes Unknown Hierarchical Not constant
BIORM [34] Multiple Reactive Autonomous No Known Clusters Not Constant
AMC [35] One Proactive Planned No Unknown Hierarchical Adaptive
MSMR [37] One Reactive Autonomous No Known Flat Constant
MSR [38] One Proactive Autonomous Yes Known Flat Adaptive
OSR [38] One Proactive Autonomous Yes Known Flat Adaptive

43039
 International Journal of Applied Engineering Research ISSN 0973-4562 Volume 10, Number 22 (2015) pp 43029-43044
© Research India Publications. http://www.ripublication.com

EASR[39] One Reactive Autonomous No Known Flat Adaptive

SMED[36] One Reactive Planned No Known Flat Adaptive
HUMS [41] One Reactive Autonomous No Known Flat Constant
ILP [42] Multiple Reactive Autonomous No Unkown Hierarchical Adaptive

Table 2 Comparision of Hierarchical Techniques for sink mobility

Scheme Location Overhead Sink Mobility Virtual Multi Applicability Effective in

Awareness cost in N/w Type Struture type sink Hot Spot problem
TTDD[54] Yes High Random Rectangular Yes Widely Strongly
MSDD [55] Yes Low Random Rectangular Yes Widely Strongly
GBEER [56] Yes Moderate Random Rectangular No Limited Moderate
VGDRA [58] Yes Low Controlled Rectanglar No Widely Strongly
Hex-DD [57] Yes Low Random Hexagonal Yes Limited N/A
MSRP [59] Yes Low Random N/A No Limited Strongly
EEMSRA [60] Yes Moderate Controlled TDMA No Rare Strongly Cluster
BVI [62] No Moderate Controlled N/A No Widely Strongly
HCDD [64] No Low Random D-Cluster No Rare Very Less
CMSE[65] No Low Random Circular No Limited Strongly
LBDD [66] Yes Low Random Line Yes Limited Very Less
PCM [68] Yes Moderate Fixed Strip No Widely Very Less
MESS [72] Yes Low Controlled Line No Limited Less
RR [69] Yes Low Random N/A Yes Widely Strongly
VLDD [73] Yes Very Low Controlled Line Yes Limited Strongly
SEAD [75] Yes High Random Steiner Tree Yes Widely Strongly
QDD [76] Yes Low Random Quad Tree Yes Limited N/A
VCCSR [77] Yes Moderate Controlled Circular No Widely Strongly

Table 3 Comparison of Non-Hierarchical techniques in mobile sink

Scheme Location Sink Mobility Multiple N/W Data Sink location update Hot Spot
Awareness Trajectory Sink overhead Aggregation Policy
DD [78] No Random No High No Global Flooding None
RR [79] Yes Random NO Low No Restricted Area Flooding None
ALURP [80] No Random No Medium No Local Area Flooding Local area only
GRAB [82] No Random No High No Global Flooding None
MG [83] No Random No Low No Restricted Area Flooding None
GCA [86] Yes Random No Low No Tree Based Flooding Near root node
RP [87] Yes Random No Low Yes Sink visit None all RP‟S
GHT [88] Yes Random No Low No Geo Hashed co-ordinates Query-data nodes
DC [89] No Random No Medium No Geometric intersect data Query data nodes
& query path
DDRP [90] No Random Yes Low No Overhearing Over hearing nodes
ER [91] Yes Random No Low No Overhearing Low possibility
GEAR [93] No Random No Low No Restricted Area None Flooding

Table 4. Types of mobility pattern

Types of Mobility Pattern Mobility Cause Speed Motion Dimensions Characteristics

Pedestrian mobility pattern Person Mobility Limited Two Dimensional Obstacle avoidance
Vehicular mobility pattern Vehicles mobility Fast One Dimensional Describe group behavior
Dynamic medium mobility pattern Fluids (Wind or water) Limited May be one, two or three Not Applicable

Network Overhead: Number of static sinks routing algorithms network overhead using mobile sink and multiple mobile
unable to gather the data for the real time application because sinks.
of large overhead. Now sink mobility is executed to overcome Quality of Services (QOS): The real time applications need
this problem. But still it needs more work to control the data on time. The music, videos and picture transmission

43040
 International Journal of Applied Engineering Research ISSN 0973-4562 Volume 10, Number 22 (2015) pp 43029-43044
© Research India Publications. http://www.ripublication.com

require best level of quality of services. The data loss removal Survey”, IEEE transactions on intelligent
and reaching of data on time both are necessary condition for transportation systems, pp 1-19.
better QOS. The energy efficient wireless sensor with best [10] Cyril Jacquemod, Benjamin Nicolle and Gilles
QOS requires more research focus. Jacquemod, (2014) “WSN for smart building
application”, Microelectronics Education (EWME),
10th European Workshop, Tallinn, pp. 102-105.
Conclusions [11] A. Cerpa, et. al, (2001) “Habitat monitoring:
In this paper, we presented various sink relocation techniques application driver for wireless communications
and sink mobility routing protocols for the wireless sensor technology”, ACM SIGCOMM, Costa Rica.
network. Sink relocation techniques are very effective in [12] M. Perillo, Z. Cheng, and W. Heinzleman, (2005)
repealing energy hole problem and network lifetime “An analysis of strategies for mitigating the sensor
enhancement. Various sink relocation techniques discussed in network hot spot problem”, in Proc. of The Second
this survey paper and energy aware methods have best International Conference on Mobile and Ubiquitous
performance in all aspects. Systems (Mobiquitous ‟05), San Diego, CA, USA.
Also routing protocols are presented for data transmission to [13] J. Li and P. Mohapatra, (2007) “Analytical modeling
mobile sink. These routing protocols are classified into and mitigation techniques for the energy hole
different categories and discussed briefly. Hierarchical problem in sensor networks”, Pervasive and Mobile
protocols are more efficient and accurate at the cost of Computing vol. 3, pp. 233-254.
increased overhead. While non-hierarchical protocols have [14] S. Olariu and I. Stojmenovic, (2006) “Design
structure less working platform which lead to low overhead. guidelines for maximizing lifetime and avoiding
But, non-hierarchical approaches provide tradeoff between energy holes in sensor networks with uniform
increased cost and accuracy. Key feature and applicability of distribution and uniform reporting”, IEEE
these protocols is highlighted. All routing protocols are INFOCOM.
compared regarding to their advantages and drawbacks. [15] A.Chakrabarti, A. Sabharwal, and B. Aazhang,
(2006) “Communication power optimization in a
sensor network with a path-constrained mobile
References observer”, ACM Trans. on Sensor Networks, vol.2,
no.3, pp. 297-324.
[1] N. Bulusu, D. Estrin, L. Girod and J. Heidemann, [16] L. Song and D. Hatzinakos, (2007) “Architecture of
(2001) “Scalable coordination for wireless sensor wireless sensor networks with mobile sinks: Sparsely
networks: self-configuring localization systems”, deployed sensors”, IEEE Trans. on Vehicular
International Symposium on Communication Theory Technology, vol. 56, pp. 1826-1836.
and Applications (ISCTA 2001), Ambleside. [17] D.Jea, A. Somasundara and M. Srivastava, (2005)
[2] J.M. Kahn, R.H. Katz and K.S.J. Pister, (1999)” Next “Multiplecontrolled mobile elements (data mules) for
century challenges: mobile networking for smart data collection in sensor networks”, Proc. 1st
dust”, Proceedings of the ACM MobiCom‟99, IEEE/ACM Intl Conf. on Distributed Computing in
Washington, USA, pp. 271-278. Sensor Systems (DCOSS), pp. 244-257.
[3] N. Noury et. al, (2000) “Monitoring behavior in [18] A. Kansal et. al, (2004) “Intelligent fluid
home using a smart fall sensor”, IEEE-EMBS infrastructure for embedded networks”, Proc. 7th
Special Topic Conference on Micro technologies in Annual Intl. Conf. on Mobile Systems, Applications
Medicine and Biology, pp. 607-610. and Services (MobiSys), pp. 111-124.
[4] J.M. Rabaey et. al, (2000) “PicoRadio supports ad [19] J. Luo et. al, (2006), “Routing towards a mobile sink
hoc ultra-low power wireless networking”, IEEE for improving lifetime in sensor networks”, 2nd
Computer Magazine (2000) pp. 42-48. IEEE/ACM Intl Conf. on Distributed Computing in
[5] B. Warneke, B. Liebowitz and K. Pister,(2001) Sensor Systems (DCOSS), pp. 480-497.
“Smart dust: communicating with a cubic millimeter [20] A.Somasundara et. al, (2006) “Controllably mobile
computer”, Computer, vol. 34, no. 1, pp. 44-51. infrastructure for low energy embedded networks”,
[6] P. Bonnet, J. Gehrke and P. Seshadri, (2000) IEEE Trans. on Mobile Computing, vol. 5, no. 8, pp.
“Querying the physical world”, IEEE personal 958-973.
communications, pp.10-15. [21] M. Wang et. al, (2005) “Exploiting sink mobility for
[7] J. Agre and L. Clare, (2000) ”An integrated maximizing sensor networks lifetime”, Proc. 38th
architecture for co-operative sensing networks „, Annual Hawaii Intl. Conf. on System Sciences, pp.
IEEE Computer Magazine, pp. 106-108. 287a-287a.
[8] M. Bhardwaj, T. Garnett and A.P. Chandrakasan, [22] S. Chellappan et. al, (2007) “Mobility limited flip-
(2001) “Upper bounds on the life time of sensor based sensor networks deployment”, IEEE
networks”, IEEE international conference on Transactions of Parallel and Distributed Systems 18
communications ICC‟01, Helsinki, Finland. (2), pp. 199-211.
[9] Hodge et. al, (2015) “Wireless Sensor Networks for [23] K. Dantu et. al, (2005) “Robomote : Enabling
Condition Monitoring in the Railway Industry: A mobility in sensor networks”, IPSN, pp. 404-409.

43041
 International Journal of Applied Engineering Research ISSN 0973-4562 Volume 10, Number 22 (2015) pp 43029-43044
© Research India Publications. http://www.ripublication.com

[24] U. Lee et. al, (2006) “Efficient data harvesting in [38] Y. Sun et. al, (2007) “Moving schemes for mobile
mobile sensor platforms”, PerCom Workshops, pp. sinks in wireless sensor networks”, in Proc. IEEE
352-356. IPCCC, pp. 101-108.
[25] C. Schindelhauer (2006), “Mobility in wireless [39] C.Wang et. al, (2014) “A network lifetime
sensor networks”, SOFSEM, pp. 100-116. enhancement method for sink relocation and it‟s
[26] Qian Dong and Waltenegus Dargie, (2013) “A analysis in wireless sensor networks”, IEEE Sensor
survey on mobility and mobility aware MAC Journal Vol.14, No. 6, pp. 1932-1943.
protocols in wireless sensor networks”, IEEE [40] S. C. Huang and R. H. Jan, (2004) “Energy-aware
Communication survey and tutorial vol. 15,NO. 1, load balanced routing schemes for sensor networks”,
pp. 88-100. in Proc. 10th Int. Conf. Parallel Distrib. Syst., pp.
[27] A. Raja and X. Su, (2009) “Mobility handling in 419-425.
MAC for wireless ad hoc networks”, Wireless [41] Y. Bi, et. al, (2007) “HUMS: An autonomous
Communications and Mobile Computing, vol. 9, pp. moving strategy for mobile sinks in data-gathering
303-311. sensor networks”, EURASIP Journal on Wireless
[28] E. W. Weisstein,(1998) The CRC concise Communications and Networking, pp. 101-108.
encyclopedia of mathematics, CRC Press. [42] S.R. Gandham et. al, (2003) “Energy efficient
[29] T. Camp, J. Boleng, and V. Davies, (2002) “ A schemes for wireless sensor networks with multiple
survey of mobility models for ad hoc network mobile base stations”, IEEE Global
research”, Wireless Communication & Mobile Telecommunications Conference, pp. 377-381.
Computing (WCMC): Special issue on Mobile Ad [43] Mohammad Hossein Anisi et. al, (2011) “ Energy-
Hoc Networking: Research, Trends and Applications, efficient data collection in wireless sensor networks”,
vol. 2, pp. 483-502. Wireless Sensor Network, pp. 329-333.
[30] M. Mart and, M. Cardei, (2005) ”Improved sensor [44] T.T. Truong, K.N. Brown and I. Sreenan,(2010)
network lifetime with multiple mobile sinks”(2009), “Using mobile sinks in wireless sensor networks to
Elsevier Journal of Pervasive and Mobile Computing improve building emergency response”, Academy
(5), pp. 542-555.. Research Colloq Uium on Wireless as an Enabling
[31] J. Luo and J. P. Hubaux, (2005) “ Joint mobility and Technology, pp. 1-4.
routing for life time elongation in wireless sensor [45] Leonidas Tzevelekas and loannis Stavrakakis, (2010)
networks”, in Proc. IEEE Inf. Commun. Conf., vol. “Sink mobility schemes for data extraction in large
3., pp. 1735-1746. scale WSNs under single or zero hop data
[32] E.M. Saad, M.H. Awadalla and R.R. Darwish,(2008) forwarding”, Wireless Conference (EW), University
“A data gathering algorithm for a mobile sink in of Athens Panepistimiopolis, Athens, Greece,
large-scale sensor networks”, The Fourth European
International Conference on Wireless and Mobile [46] Junzhao Du et. al, (2012) “ Rendezvous data
Communications, pp. 207-213. collection using a mobile element in heterogeneous
[33] M. Marta and M. Cardei, (2009) “Improved sensor sensor networks”, International Journal of
network lifetime with multiple mobile sinks”, J. Distributed Sensor Networks 12 pages, Article 10
Pervas. Mobile Computing, vol. 5, no. 5, pp. 542- 686172, DOl:10.1155/2012/686172.
555. [47] R.G. Raj and V. Balakrishnan, (2011) “A model for
[34] Sumit Kataria and Aarti Jain,(2013) “ Bio inspired determining the degree of contradictions in
optimal relocation of mobile sink nodes in wireless information”, Malaysian Journal of Computer
sensor networks”, Emerging Trends in Science, pp. 160-167.
Communication, Control, Signal Processing & [48] Zhou Sha et. al, (2010) “An anti-detection moving
Computing Applications (C2SPCA), International strategy for mobile sink”, IEEE Proceeding of Global
Conference, pp. 1-6. Telecommunications Conference, pp. 1-5.
[35] A.A. Somasundara et. al, (2006) “Controllably [49] Long Cheng et. al, (2009) “Query-based data
mobile infrastructure for low energy embedded collection in wireless sensor networks with mobile
networks”, IEEE Transactions on Mobile sinks”, International Conference on Wireless
Computing, pp. 958-973. Communications and Mobile Computing:
[36] Bin wang, Dongliang Xie, Canfeng Chen, Jian Ma, Connecting the World Wirelessly, pp. 1157-1162.
and Shiduan Chen, (2008) “Employing mobile sink [50] Athanasios Kinalis et. al, (2012) “ Biased sink
in event-driven wireless sensor networks”, IEEE mobility with adaptive stop times for low latency
communication society proceedings in ICC., pp. 188- data collection in sensor networks”, STREP ICT-
192. 257466, information Fusion Journal, pp. 1566-2535.
[37] L. Sun, Y. Bi, and J. Ma, (2006) “ A moving strategy [51] Taisoo Park et. al, (2009) “Energy efficient and
for mobile sinks in wireless sensor networks”, in seamless data collection with mobile sinks in
Proc. 2nd IEEE Workshop Wireless Mesh Netw., pp. massive sensor networks”, Parallel and Distributed
151-153. Processing, pp. 1-8.

43042
 International Journal of Applied Engineering Research ISSN 0973-4562 Volume 10, Number 22 (2015) pp 43029-43044
© Research India Publications. http://www.ripublication.com

[52] Kansal et. al, (2006) “Intelligent fluid infrastructure networks”, Int. J. Distrib. Sensor Netw., vol., pp. 1-
for embedded networks”, ACM MobiSYS‟04, 15.
Boston, Massachusetts, USA. [65] Wei-Cheng Chu, kuo-Feng Ssu, (2014) “Sink
[53] Zhao W., Ammar, M and Zegura. E.,(2004) “A discovery in location-free and Mobile sink wireless
message ferrying approach for data delivery in sparse sensor networks”, Elsevier, Computer networks, pp.
mobile ad hoc network”, Proceedings of the 5th 123-140.
ACM international symposium on Mobile ad hoc [66] E. Ben Hamida and G. Chelius,(2005) “A line-based
networking and computing, pp. 187-198. data dissemination protocol for wireless sensor
54] H. Luo et. al, (2005) “ TTDD: Two-tier data networks with mobile sink”, in IEEE Int. Conf. on
dissemination in large-scale wireless sensor Communications, 2008. ICC ‟08, pp. 2201-2205.
networks”, Wireless Networks, vol. 11, nos. 1-2, pp. [67] J.-H. Shin et. al,(2005) “Railroad: Virtual
161-175. infrastructure for data dissemination in wireless
[55] XIE Dongliang et. al, (2014) “ Multiple mobile sink sensor networks”, in Proc. 2nd ACM int. workshop
data dissemination mechanism for large scale on Performance evaluation of wireless ad hoc,
wireless sensor network”, China communication sensor, and ubiquitous networks. PE-WASUN ‟05,
Supplement No 1, Vol. 11, issue 13, pp. 1-8. pp. 168-174.
[56] K. Kweon et. al, (2009) “Grid-based energy-efficient [68] Shuai Gao, Hongke Zhang and Sajal Das, (2009)
routing from multiple sources to multiple mobile “Efficient data collection in wireless sensor networks
sinks in wireless sensor networks”, in Wireless with path-constrained mobile sinks”, World of
Pervasive Computing, 2009. ISWPC 2009, 4th Int. Wireless, Mobile and Multimedia Networks &
Symp., pp. 1 -5. Workshops, International Symposium , pp. 1-9.
[57] A. Erman, A. Dilo, and P. Havinga, (2012) “A virtual [69] C. Tunca, et.al, (2012), “Ring routing: An energy-
infrastructure based on honeycomb tessellation for efficient routing protocol for wireless sensor
data dissemination in multi-sink mobile wireless networks with a mobile sink”, in Signal Processing
sensor networks”, EURASIP J. on Wireless and Communications Applications Conf. (SIU), pp.
Communications and Networking, vol. 2012, no. 1, 1-4.
p. 1-17. [70] R. C. Shah et.al, (2003) “Data MULEs: Modeling a
[58] Abdul Waheed Khan et. al, (2015) “ VGDRA: A three-tier architecture for sparse sensor networks”,
virtual grid-based dynamic routes adjustment scheme Proceedings of IEEE SNPA, pp. 30-41.
for mobile sink-based wireless sensor networks”, [71] David Jea et. al, (2005) “Multiple controlled mobile
IEEE Sensors Journal, vol. 15, no. 1. pp. 526-534. elements (data mules) for data collection in sensor
[59] B. Nazir and H. Hasbullah, (2010) “Mobile sink networks”, Springer-Verlag Berlin Heidelberg,
based routing protocol (MSRP) for prolonging DCOSS 2005, LNCS 3560, pp. 244-257.
network lifetime in clustered wireless sensor [72] B. Tang et. al, (2012) “A novel data retrieving
network”, in Computer Applications and Industrial mechanism in wireless sensor networks with path-
Electronics (ICCAIE), Int. Conference, pp. 624 -629. limited mobile sink”, Int. J. Grid Distrib. Comput.,
[60] Y. Xun-Xin and Z. Rui-Hua,(2011) “ An energy- vol. 5, no. 3, pp. 133-140.
efficient mobile sink routing algorithm for wireless [73] Hee-Sook Mo et. al, (2013) “Virtual Line-Based
sensor networks”, in Wireless Communications, Data Dissemination for Mobile Sink Groups in
Networking and Mobile Computing (WiCOM), Wireless Sensor Networks”, IEEE Communications
(2011) 7th Int. Conf., pp. 1 -4. Letters, vol. 17, no. 9, pp. 1864-1867.
[61] W. R. Heinzelman, A. Chandrakasan, and H. [74] S. Park et. al, (2010) “ Mobile geocasting to support
Balakrishnan,(2000) “Energy efficient mobile sink groups in wireless sensor networks”,
communication protocol for wireless microsensor IEEE Commun. Lett., vol. 14, no. 10, pp. 939-941.
networks”, in Proc. 33rd Hawaii Int. Conf. on [75] H. S. Kim et. al, (2003) “ Dissemination: Minimum
System Sciences, ser. HICSS ‟00, vol. 8. energy asynchronous dissemination to mobile sinks
Washington, DC, USA: IEEE Computer Society, pp. in wireless sensor networks”, Proceedings of ACM
8020-8030. SenSys, November 2003.
[62] S. Oh, E. Lee, S. Park, J. Jung, and S.-H. Kim, [76] Z. Mir and Y.-B. Ko, (2007) “A quad-tree-based
(2010) “Communication scheme to support sink hierarchical data dissemination for mobile sensor
mobility in multi-hop clustered wireless sensor networks”, Telecommunication Systems, vol. 36, pp.
networks ”, in Proc. 24th IEEE Int. Conf. Adv. Inf. 117-128.
Netw., pp. 866-872. [77] T.-S. Chen et. al, (2013) “Geographic converge cast
[63] O. Younis and S. Fahmy, (2004) “ HEED: A hybrid, using mobile sink in wireless sensor networks”,
energy-efficient, distributed clustering approach for Comput.Commun., vol. 36, no. 4, pp. 445-458.
ad hoc sensor networks”, IEEE Trans. Mobile [78] I. Chalermek et. al, (2000) „Directed Diffusion: A
Comput., vol. 3, no. 4, pp. 366-379. scalable and robust communication paradigm for
[64] L. Buttyán and P. Schaffer, (2010) “Position-based sensor networks‟, in Proc. ACM SIGMOBILE Int.
aggregator node election in wireless sensor Conf. Mobile Comput. Netw. (MOBICOM), pp. 56-
67.

43043
 International Journal of Applied Engineering Research ISSN 0973-4562 Volume 10, Number 22 (2015) pp 43029-43044
© Research India Publications. http://www.ripublication.com

[79] D. Braginsky and D. Estrin,(2002) “Rumor Routing Author’s Biography

algorithm for sensor networks”, proc. 1st Wksp.
Sensor networks and App.,Atlanta, GA., pp. 22-31.
[80] G. Wang et. al, (2009) „Adaptive location updates for
mobile sinks in wireless sensor networks‟, The
journal of supercomputing 47 (2), pp. 127-145.
[81] G. wang et. al, (2005) “Local update-based routing
protocol in wireless sensor networks with mobile
sinks”, IEEE Communications Society (ICC), pp.
3094-3099. Er. Vicky kumar is pursuing his Ph.D from National Institute
[82] F. Yen et. al, (2005) “Gradient broadcast: A robust of Technology, Hamirpur, India. He has passed his M.Tech
data delivery protocol for large scale sensor from LPU Jalandhar. His research interest includes wireless
networks”, Wireless Networks, vol. 11, pp. 285-298, sensor networks and routing algorithms.
2005.
[83] S. Park et. al, (2010) “ Mobile geocasting to support
mobile sink groups in wireless sensor networks”,
IEEE Commun. Lett., vol. 14, no. 10, pp. 939-941.
[84] I. Stojmenovic, (2004) “Geocasting with guaranteed
delivery in sensor networks”, IEEE Wireless
Commun., vol. 11, no. 6, pp. 29-37.
[85] S. Ni et. al, (1999) „The broadcast storm problem in a
mobile ad hoc network‟, in Proc. 1999 ACM Dr. Ashok Kumar received his Ph.D. degree from NIT
MOBICOM, pp. 151-162. Hamirpur in the field of Wireless Sensor Networks. He is
[86] X. Chen and M. Xu,(2003) “A geographical cellular- working as Associate Professor in the Department of
like architecture for wireless sensor networks”, in Electronics & Communication Engineering, NIT Hamirpur,
Proc. Mobile Ad-Hoc Sensor Netw., pp. 249-258. Himachal Pradesh, India. He has published more than 40
[87] Guoliang Xing et. al, (2012) “Efficient rendezvous research papers in International/National journals and
algorithms for mobility-enabled wireless sensor Conferences in the area of Wireless Communication and
networks”, IEEE Transactions on Mobile Wireless sensor networks.
Computing, vol. 11, no. 1, pp. 47-60.
[88] S. Ratnasamy et. al, (2002) “GHT: A geographic
hash table for data-centric storage”, in Proc. of the
1st ACM int. workshop on Wireless sensor networks
and applications, ser. WSNA ‟02. New York, NY,
USA: ACM, pp. 78-87.
[89] G. Shi et. al, (2012) “Double-blind data discovery
using double cross for large-scale wireless sensor
networks with mobile sinks”, IEEE Trans. Veh. Er. Gaurav is pursuing his Ph.D from National Institute of
Technol., vol. 61, no. 5, pp. 2294 -2304. Technology, Hamirpur, India. He has passed his M.Tech from
[90] L. Shi, B. Zhang, H. T. Mouftah, and J. Ma, (2012) Kurukshetra University. His research interest includes
“DDRP: An efficient data driven routing protocol for wireless sensor networks and routing algorithms.
wireless sensor networks with mobile sinks”, Int. J.
Communication Systems.
[91] F. Yu, S. Park, E. Lee, and S.-H. Kim, (2010)
“Elastic routing: A novel geographic routing for
mobile sinks in wireless sensor networks”,
Communications, IET, vol. 4, no. 6, pp. 716 -727.
[92] Karp B. and Kung H.T., (2000) “GPSR: greedy
perimeter stateless routing for wireless networks”,
Proc. Sixth Annual Int. Conf. Mobile Computing and
Networking, pp. 243-254
[93] Y.Yu, D. Estrin and R.Govindan, (2001)
“Geographical and Energy-Aware Routing: A
recursive data dissemination protocol for wireless
sensor networks”, UCLA Comp. Sci. Dept. tech. rep.
UCLA-CSD TR-010023.

43044