Path constrained data gathering scheme for wireless sensor networks with mobile elements

International Journal of Wireless & Mobile Networks (IJWMN) Vol. 6, No. 3, June 2014
DOI : 10.5121/ijwmn.2014.6305 69
PATH-CONSTRAINED DATA GATHERING SCHEME
FOR WIRELESS SENSOR NETWORKS WITH MOBILE
ELEMENTS
Bassam A. Alqaralleh and Khaled Almi’ani
Al-Hussein Bin Talal University, Jordan
ABSTRACT
Wireless Sensor Networks (WSNs) have emerged as a promising solution for variety of applications.
Recently, in order to increase the lifetime of the network, many proposals have introduced the use of
Mobile Elements (MEs) as a mechanical carrier to collect data. In this paper, we investigate the problem of
designing the mobile element tour to visit subset of the nodes, termed as caching points, where the length of
the mobile element tour is bounded by pre-determined length. Caching can be implemented at various
points on the network such that any node in the network is at most k-hops away from one of these caching
points. To address this problem, we present heuristic-based solution. Our solution works by partitioning
the network such that the depth of each partition is bounded by k. Then, in each partition, the minimum
number of required caching points is identified. We compare the resulting performance of our algorithm
with the best known comparable schemes in the literature.
KEYWORDS
Wireless Sensor Networks, Mobile Element
1. INTRODUCTION
Wireless Sensor Networks (WSNs) composed of multifunctional miniature devices with sensing,
computation and wireless communication capabilities. Such devices are normally battery
operated. Thus, recharging a sensor node is impractical because they are typically designed for
hostile environments; therefore, energy efficiency is critical. In typical static multi-hop data
communication paradigm, the sensors around the sink are likely the first to run out of energy.
This is due to the fact that these sensors carry heavier traffic loads. Once these sensor nodes fail,
the operational lifetime of the networks ends and the network stops working because the entire
network becomes unable to communicate with the sink. Therefore, reducing energy consumption
in WSNs is becoming a major design challenge.
In order to significantly increase the lifetime of the network via considerably reducing the energy
consumption, many proposals [1][2][3] have introduced the use of the Mobile Elements (MEs) as
an efficient approach. A mobile element roams through the network and collects data from
sensor nodes via short range (single-hop) communications. Therefore, in comparison with multi-
hop communications the energy consumption can be considerably reduced. However, the speed
of the mobile element is typically low[4][5], and therefore reducing the data gathering latency is
important to ensure the efficiency of such approach.
To address this problem, several proposals presented a hybrid approach, which combines multi-
hop forwarding with the use of mobile elements. In this approach, each mobile element visits
subset of the nodes termed as caching points. These caching points store the data of the nodes

70
that are not included in the tour of the mobile element. Once a mobile element becomes within
the transmission range of a caching point, the caching point transmits its data to the mobile
element. By adopting such an approach, the mobile element gathers the data of the entire network
without the need for visiting each node physically.
In this paper, we focus on the problem of reducing the latency of the mobile elements tours.
Accordingly, we investigate the K-Hop tour planning (KH-tour) problem. This problem deals
with designing the shortest possible tour for the mobile, such that the tour start and end at the
sink, and the maximum number of hops between any node and the tour is bounded by pre-
determined value.
To address this problem, we present an algorithmic-based solution. This solution is based on
partitioning the network into groups of partitions, where the depth of each partition is selected to
ensure that one node in each partition can be identified as a caching point without violating the
hop constraints. Then, the tour of the mobile element will be constructed to visit all nodes that
have been identified as caching points.
The rest of the paper is organized as follows. Section 2 provides formal definition of the
presented problem. Section 3 presents the related work in this area. Section 4 presents the
heuristic solution of the presented problem. In Section 6, the evaluation for the proposed
heuristics is presented. Finally, Section 7 concludes the paper.
2. RELATED WORKS
Many proposals in the recent literature have studied the use of mobile elements to reduce the
energy consumption in sensor networks. In this section, we discuss the most related work in the
literature.
The problem presented in our work can be recognized as an extended version of the tour
scheduling problem presented by Somasundara et al. [5][6]. In this tour scheduling problem, the
authors assume that each node must be visited within a time-deadline to avoid buffer overflow.
Accordingly, the authors proposed different heuristics to determine the path of the mobile
elements.
Guney et al. [7] formulated the problem of finding the optimal sink trajectory and the data flow
routs as mixed integer programming formulations. Accordingly, they presented several heuristics
to address this problem. Liang et al. [10]also presented a mixed integer programming formulation
for similar problem, where they incorporate the constraint of the travelling to the formulation.
In [9], the authors proposed a data collection scheme aims to increase the network throughput. In
this scheme, the mobile element is assumed to move periodically in pre-determined path, which is
expected to visit a subset of the nodes called subsinks. The main goal of this scheme is optimizing
the assignment of the sensor nodes to subsinks in order to increase the network throughput.
Zhao et al [11][12]investigated the problem of maximizing the overall network utility. Therefore,
they presented two distributed algorithms for data gathering where the mobile sink stays at each
anchor point (gathering point) for a period of sojourn time and collects data from nearby sensors
via multi-hop communications. They considered both variable and fixed sojourn time.
The problem presented in this paper share some similarities with the Vehicle Routing Problem
(VRP) [13]. Given a fleet of vehicles assigned to a depot, VRP deals with determining the fleet
routes to deliver goods from a depot to customers while minimizing the vehicles' total travel cost.

71
Among the vehicle routing problem variations, the Vehicle Routing Problem with Time Windows
(VRPTW) [14]is the closest to our problem. In VRPTW, each customer must be visited by
exactly one vehicle and within a pre-defined time interval. The problem presented in this work
also share some similarities with the Deadline Travelling Salesman Problem (Deadline-TSP)[14],
which seeks the minimum tour length for a salesman to visit a set of cities, where each city must
be visited before a pre-determined time deadline. In particular, the special case where all cites
have the same deadline reduces Deadline-TSP to the well-known orienteering problem (OP) [15];
thus, our problem can be considered as a generalization of the orienteering problem.
The problem presented in this work shares some similarities with the problem proposed by Ma et
al. [16]. In their proposal, they explored two settings where each polling point is either the
location of a sensor or a point in the network area. On the other hand, the mobile element can
communicate with one or more identified nodes in order to collect their data. Their goal is to
design the mobile element tour which consists of polling points in order to find the shortest
possible tour where each node is either on the tour or one-hop away from the tour.
The problem presented in this share some similarities with minimum-energy Rendezvous
Planning Problem (RPP) [17][18]. In this problem, the objective is to determine the mobile
element path such the total Euclidean distance between nodes not included in the tour and the tour
is minimized. In [17], the authors presented the Rendezvous Design for Variable Tracks (RD-VT)
algorithm. The process of this algorithm starts by construction the Steiner Minimum Tree (SMT)
that connects the source nodes. Then, the obtained tree will be traversed in pre-order until no
more nodes can be visited without violating the deadline constraint. In this algorithm the visited
nodes is identified as the caching points. Xing et al. [3] provided a utility-based algorithm and
address the optimal case for restricted version of the problem. Many proposals [19] have also
investigated this problem.
3. PROBLEM DEFINITION
An instance of the KH-tour problem consists of an undirected complete graph ‫ܩ‬ ൌ ‫,ܸۃ‬ ‫,ۄܧ‬ ‫ݒ‬௦. ܸ
is the set that represents the location of the sensor nodes in the network.‫ݒ‬௦ represents the
location of the sink. ‫ܧ‬ is the set of edges that represents the communication pattern in the
network. For each pair of nodes ‫ݒ‬௜, ‫ݒ‬௝ ‫א‬ ܸ there is an edge between these two nodes, if they are
within each other communication range. ݀ሺ‫ݒ‬௜, ‫ݒ‬௝ሻ represents the Euclidean distance that the
mobile element will travel between nodes ‫ݒ‬௜ and ‫ݒ‬௝. In addition, we use a pre-determined value ݇
that represents the maximum number of hops allowed between any node and the tour.
A solution to the KH-tour problem consists of a tour (a path in ‫)ܩ‬ that starts and ends in ‫ݒ‬௦,
where the objective is to minimize the Euclidean distance of this tour, such that each node ‫ݒ‬௜ ‫א‬ ܸ
is at most k-hops away from the tour.
4. INTEGER LINEAR PROGRAM FORMULATION
Let ܸ ൌ ሼ‫ݒ‬଴, ‫ݒ‬ଵ, … , ‫ݒ‬௡ሽ be the set the represents the nodes in the networkሺ‫ݒ‬௦ ‫א‬ ܸሻ. Also, let
݀ሺ‫,ݑ‬ ‫ݒ‬ሻ be the distance which the mobile element needs to travel to reach node ‫ݒ‬ from node ‫.ݑ‬
We use the ‫ݕ‬௨,௩ as an (0-1) indicator for each pair of nodes ሺ‫,ݑ‬ ‫ݒ‬ሻ ‫א‬ ܸ such that ‫ݕ‬௨,௩ ൌ 1, if the
mobile element travels between node ‫ݒ‬ and node ‫,ݑ‬ and 0 otherwise. Also, let ‫ݖ‬௨be an integer
variable for each node ‫ݑ‬ ‫א‬ ܸ, taking only positive values, showing the order in which the nodes
are visited in the resulting tour. In addition, let ‫ݔ‬௨,௩be an integer variable for each pair of nodes
‫,ݒ‬ ‫ݑ‬ ‫א‬ ܸ such that ‫ݔ‬௨,௩ ൌ 1, if nodes ‫ݒ‬ and ‫ݑ‬ are one hop away from each other, and node ‫ݑ‬

72
stores node ‫ݒ‬ data; 0 otherwise. Also, let ݄ሺ‫,ݒ‬ ‫ݑ‬ሻ be the number of hops between nodes ‫ݒ‬ and ‫.ݑ‬
Given the Integer Linear Program for our problem can be given as follows:
Minimize ෍ ‫ݕ‬௜,௝
௜,௝‫א‬௏
· ݀ሺ݅, ݆ሻ
(1)
Subject to:
෍ ‫ݕ‬௝,௜ ൑ 1
௜‫א‬௏
‫݆׊‬ ‫א‬ ܸ(2)
෍ ‫ݕ‬௜,௝ ൑ 1
௜‫א‬௏
‫݆׊‬ ‫א‬ ܸ(3)
෍ ‫ݕ‬௜,௝ െ ෍ ‫ݕ‬௝,௜ ൌ 0
௜‫א‬௏௜‫א‬௏
‫݅׊‬ ‫א‬ ܸ(4)
‫ݔ‬௜,௝ · ݄൫‫ݒ‬௜, ‫ݒ‬௝൯ ൑ ݇ ‫,݅׊‬ ݆ ‫א‬ ܸ(5)
෍ ‫ݔ‬௜,௝
௜‫א‬௏
൅ ෍ ‫ݕ‬௝,௜
௜‫א‬௏
ൌ 1
‫݆׊‬ ‫א‬ ܸ(6)
‫ݖ‬௜ െ ‫ݖ‬௝ ൅ ݊ · ‫ݕ‬௜,௝ ൑ ݊ െ 1 ‫݆׊‬ ‫א‬ ܸ ሼ‫ݒ‬௦ሽ⁄ (7)
‫݅׊‬ ‫א‬ ܸ
Constraints (2-4) ensure that the load at each node is balanced. Constraint (5) ensures that the
number of hops between any node and the tour is at most. Each node must be either involved in
the mobile element tour or connected to a node involved in this tour, and this is represented by
constraint (6). Constraint (7) is the sub-tour elimination constraint. Typically, it is hard to solve
such problem. However, the ILP is given to take close look at the problem.
5. ALGORITHMIC-BASED SOLUTION
Our objective is to determine the shortest mobile element tour such that the number of hops
between any node not included in the tour and the tour is at most ݇. Towards this end, in this
section, we present the Graph Partitioning (GP) algorithm. The main idea of this algorithm is to
divide the graph of the network into partitions, such that the depth of each partition is at most
2 · ݇ hops. As we will discuss in this section, this algorithm is designed to address the situation
where the nodes are uniformly deployed, and the partitioning process aims to simplify the step of
identifying the caching points.
The Graph Partitioning (GP) algorithm starts by identifying the center node of the graph. This
node is the closest to all other nodes (in term of number of hops) inside the graph. Then the
process iterates to identify the nodes belong to each partition. Once the nodes in each partition are
identified, they are flagged by their partition number, to make sure that they will not be
reconsidered during the partitioning steps. Then, the caching point identification step starts the
process of identifying the caching points in each partition. This process aims to identify the
minimum number of required caching points, such that the number of hops between any node and
one of these nodes is at most ݇. As we will see next, this process works by establishing a path
from the nodes, which have the highest number of neighbors. Two nodes are neighbors, if they
are at most k-hops away from each other. Establishing this path aims to select the minimum

73
number of required caching points. The last step of this algorithm is to construct the mobile
element tour from the identified caching points. Algorithm 1 illustrates these steps. Now, we
discuss the steps of this algorithm in more details.
Algorithm 1. The Graph Partitioning (GP) algorithm
1: procedure‫ܲܩ‬ሺ‫,ܩ‬ ‫ܩ‬ᇱ
, ݇ሻ
Inputs: the graphs ‫ܩ‬ and ‫’ܩ‬
݇, maximum hops constraints
Outputs: A tour (ܶ) and forwarding trees (ܴ)
2: ܿ݊ ՚CenterNode(‫)ܩ‬
3: ܲ ՚Partition(‫,ܩ‬ ܿ݊)
4: ‫ܥ‬ ՚CachingPoints(ܲ, ‫)ܩ‬
5:ܶ ՚TourBuild(C)
6:ܴ ՚BuildRouting(ܶ, ‫)ܩ‬
7: procedure CenterNode (‫)ܩ‬
8: for each vertex ‫ݒ‬ in ‫ܩ‬do
9: dist(‫)ݒ‬ ← sum of hop-distances from all nodes
10: end for
11: return node with minimum dist(v)
12: end procedure
13: procedure Partition (‫,ܩ‬ ܿ݊)
14: for each vertex ‫ݒ‬ in ‫ܩ‬do
15: hops(‫)ݒ‬ ←number of hops in the shortest path between ‫ݒ‬ and ܿ݊
16: ݅ ՚ ‫ہ‬hopsሺ‫ݒ‬ሻ 2 · ݇⁄ ‫1+ۂ‬
17: assign‫ݒ‬ to ܲ௜
18: end for
19: returnܲ
20:end procedure
21: procedure CachingPoints (‫,ܩ‬ ܲ, ݇)
22: for each ܲ௜ in ܲdo
23: for each ‫ݒ‬௜, ‫ݒ‬௝ in ܲ௜do
24: if path between ‫ݒ‬௜and‫ݒ‬௝ at most k hops do
25: add edge between ‫ݒ‬௜and‫ݒ‬௝ in ‫ܩ‬
26: end for
27: end for
28: for each ܲ௜ in ܲdo
29: add node with highest degree to ‫ܥ‬௜
31: Remove this node and its one-hope neighbors from ܲ௜
32: end for
33: returnܲ
34:end procedure
5.1 The partitioning step
This step starts by identifying the center node of the graph. Then, starting by the first partition, the
process iterate to identify the nodes belong to each partition. The identity of the nodes that are
assigned to each partition is selected based on the partition number, and the number of hops
between any node and center node. For partition ݅, the set of the nodes that belong to this partition
consists of the nodes that are not assigned to any previous partition, and can reach the center node
within ݅ · 2 · ݇ hops. As we can see, by performing such partitioning process, the network will be
divided into partitions, where the depth of each partition is 2 · ݇. Now, in each partition there

74
must be a set of nodes that can reach all other nodes inside the partition within ݇-hops. By
performing this partitioning step, we attempt to reduce the solution space for selecting the caching
points in an efficient manner in order to simplify the process of identifying the caching nodes.
5.2 The caching point identification step
In this step, the process iterates to determine the caching points for each partition. Starting at
partition number1, for each pair of nodes ‫ݒ‬௜, ‫ݒ‬௝ ‫א‬ ܸ belong to this partition, we add an edge
between these two nodes to ‫,ܧ‬ if they are k-hops away from each other. Only the nodes that
belong to the same partition will be considered during the process of determining the k-hops
connectivity. Once the k-hops connectivity for all partitioned are examined, the process proceed
by constructing a set of the caching points for each partition ‫ܥ‬ ൌ ሾ‫ܥ‬ଵ, ‫ܥ‬ଶ, … , ‫ܥ‬௜ሿ. In each
partition(݅), based on the degree of the nodes, subset of the partition nodes will be added to ‫ܥ‬௜. In
each iteration, the node with the highest degree will be add to ‫ܥ‬௜. This node and all of its one-hop
neighbors (inside the partition) will be flagged to ensure that they are considered in consecutive
iterations. This process stops when all nodes inside each partition are flagged. Now, the set of the
caching points consists of the collection of caching points in all partitions.
5.3 The routing trees construction and tour building steps
Once the caching points set are identified, each node not included in this set will be assigned to its
nearest caching point. Then, for each caching point and the nodes which are assigned to this
caching point, a Minimum Spanning Tree (MST) is created to establish multi-hop forwarding
trees. The tour of the mobile element consists of the caching points and the sink. This is
established using Christofides algorithm [18].
6. EXPERIMENTAL EVALUATION
To evaluate the presented algorithm’s performance, we conducted an extensive set of experiments
using the J-sim simulator [19]. the area of the network is 250,000m2
, and The radio parameters
are set according to the MICAz data sheet [20], namely: the radio bandwidth is 250 Kbps, the
transmission power is 21 mW, the receiving power is 15 mW, and the initial battery power is 10
Joules. The packet has a fixed size of 100 bytes. Each experiment is taken over an average of 10
different realizations of random topologies. We are particularly interested in investigating the
following metrics:
• The lifetime of the network
• Number of caching points.
The parameters we consider in our experiments look at varying the number of nodes. We
consider the following deployment scenarios:
• Uniform density deployment: in this scenario, we assume that the nodes are uniformly
deployed in a square area of 500× 500m2
.
• Variable density deployment: in this scenario, we divide the network into a 10×10 grid
of squares, where each square is 50×50m2
. We randomly choose 30 of the squares, and
in each one of those squares we fix the node density to be x times the density in the
remaining squares. x is a density parameter, which in most experiments (unless
mentioned otherwise) is set to x = 5.

75
To benchmark the presented algorithm performance, we compared it against the Spanning Tree
Covering Algorithm, which is proposed by Ma et al.[16], we will refer to this algorithm as the T-
Covering algorithm. This algorithm starts by selecting the sink as the initial point in the mobile
element tour. Then, in each iteration and based on a cost function, the polling point with the
lowest cost will be added to the tour. Ma et al.[18] defines the polling point as a point in the
network, where the mobile element can communicate with one or more sensor nodes via a single
hop transmission. For a given polling point, this cost function returns the shortest distance
between this polling point and a node belong to the tour, divided by the number of new nodes will
be covered if this polling point is selected. This process iterate until all nodes are covered by the
tour. To ensure the fairness of the comparison, we restrict the
Figure 1: Total travelling time against the number ofnodes, for the uniform density deployment
scenario
Figure 2: Total travelling time against the number of nodes, for the variable density deployment
scenario
polling points in the T-covering algorithm to be the actual sensor nodes, and in the T-covering
algorithm, we add a direct edge between any two nodes, if they are k-hops from each other.
First, we evaluate the impact of the number of nodes on the number of tours each algorithm
obtains. Figure 1 and Figure 2 show the results for deployment scenarios. From the figures, we
can see that in the uniform deployment scenario, the T-Covering algorithm outperforms the GP
algorithm, where in the variable deployment scenario, the GP algorithm outperform the T-
Covering algorithm. To understand this behavior, let us clarify the relationship between the
deployment scenario and the expected performance for each algorithm. In the T-Covering
algorithm, the employed cost function selects the node with the lowest cost to be included in the
8000
10000
12000
14000
16000
18000
20000
600 700 800 900 1000
Totaltravellingtime
Number of nodes
T-Covering
GB
8000
10000
12000
14000
16000
18000
20000
22000
600 700 800 900 1000
Totaltravellingtime
Number of nodes
T-Covering
GP

76
tour. The cost value for each node is obtained by dividing the distance between this node and the
tour, by the number of nodes which will be covered if this node is included in the tour. Such cost
function results in degrading
Figure 3: Average number of caching points against the number of nodes, for the uniform density
deployment scenario
Figure 4: Average number of caching points against the number of nodes, for the variable density
deployment scenario
the performance of the T-Covering algorithm in the variable deployment scenario compared to the
uniform deployment scenario. This occurs because in the uniform deployment scenario, we
expect that the T-Covering algorithm will select caching points uniformly located across the
entire network. Furthermore, in the variable deployment scenario, this cost function is expected to
result in selecting caching points mainly in the dense areas in the network and this is expected to
increase the length of the obtained tour. In the GP algorithm, the process of partitioning the
network results in disconnecting the process of selecting the caching points from the deployment
distribution. These are the main factors which explain the performance of our algorithm.
We proceed to investigate the impact of the number of nodes on the number of caching points that
each algorithm obtains. Figures 3 and 4 show the results for both the uniform density and the
variable density deployment scenarios; respectively. From the figures, we can see that in the
uniformly deployment scenario, the T-Covering algorithm obtains slightly lower number of
caching points compared to the GP algorithm. On the other hand, in variable deployment
scenario, the GP algorithm obtains significantly lower number of caching points. This is also due
to the factors mentioned above, since in the variable deployment scenario, we expect that the T-
covering algorithm obtains longer tour compared to the GP algorithm.
0
20
40
60
80
100
120
600 700 800 900 1000
AVGnumberofCPs
Number of nodes
T-Covering
GP
0
20
40
60
80
100
120
600 700 800 900 1000
AVGnumberofCPs
Number of nodes
T-Covering
GP

77
7. CONCLUSIONS
In this work, we address the problem of designing the mobile element tour, such that the length of
this tour is below a pre-determined value, and each node is at most k-hops away from the tour.
Accordingly, we propose a heuristic-based solution which is based on dividing the network into
partitions, with pre-determined depth. Then, a subset of the nodes in each partition which will act
as caching points for data must be identified such that the length of the tour that consists of these
caching points and the sink node is below L.
An interesting open problem is considering some application scenarios where the latency
requirements of data gathering may vary from one network’s location to another. For example,
some areas in the network need to be visited more frequently than others. In this case the tour
length constraints would be different for different areas.
REFERENCES
[1] Y. Gu, D. Bozdag, E. Ekici, F. Ozguner, and C. G. Lee, “Partitioning based mobile element
scheduling in wireless sensor networks,” in Sensor and Ad Hoc Communications and Networks,
2005. IEEE SECON 2005. 2005 Second Annual IEEE Communications Society Conference on, 2005,
pp. 386–395.
[2] G. Xing, T. Wang, Z. Xie, and W. Jia, “Rendezvous planning in wireless sensor networks with mobile
elements,” IEEE Trans. Mob. Comput., vol. 7, no. 12, pp. 1430–1443, Dec. 2008.
[3] K. Dantu, M. Rahimi, H. Shah, S. Babel, A. Dhariwal, and G. S. Sukhatme, “Robomote: enabling
mobility in sensor networks,” in Proceedings of the 4th international symposium on Information
processing in sensor networks (IPSN), 2005, pp. 404 – 409.
[4] R. Pon, M. A. Batalin, J. Gordon, A. Kansal, D. Liu, M. Rahimi, L. Shirachi, Y. Yu, M. Hansen, W. J.
Kaiser, and others, “Networked infomechanical systems: a mobile embedded networked sensor
platform,” in Proceedings of the 4th international symposium on Information processing in sensor
networks, 2005, pp. 376 – 381.
[5] A. Somasundara, A. Ramamoorthy, and B. Srivastava, “Mobile Element Scheduling for Efficient
Data Collection in Wireless Sensor Networks with Dynamic Deadlines,” in Real-Time Systems
Symposium, 2004. Proceedings. 25th IEEE International, 2004, pp. 296 – 305.
[6] A. A. Somasundara, A. Ramamoorthy, and M. B. Srivastava, “Mobile element scheduling with
dynamic deadlines,” IEEE Trans. Mob. Comput., vol. 6, no. 4, pp. 395–410, 2007.
[7] E. Güney, I. K. Altmel, N. Aras, and C. Ersoy, “Efficient integer programming formulations for
optimum sink location and routing in wireless sensor networks,” in Proceedings of the 23rd
International Symposium on Computer and Information Sciences, 2008, pp. 1–6.
[8] W. Liang, J. Luo, and X. Xu, “Prolonging Network Lifetime via a Controlled Mobile Sink in Wireless
Sensor Networks,” in Proceedings of Global Telecommunications Conference (GLOBECOM 2010),
2010, pp. 1–6.
[9] B. Hull, V. Bychkovsky, Y. Zhang, K. Chen, M. Goraczko, A. Miu, E. Shih, H. Balakrishnan, and S.
Madden, “CarTel: a distributed mobile sensor computing system,” in Proceedings of the 4th
international conference on Embedded networked sensor systems, 2006, pp. 1–14.
[10] M. Zhao and Y. Yang, “Optimization-Based Distributed Algorithms for Mobile Data Gathering in
Wireless Sensor Networks,” IEEE Trans. Mob. Comput., vol. 11, no. 10, pp. 1464–1477, 2012.
[11] M. Zhao and Y. Yang, “efﬁcient data gathering with mobile collectors and space-division multiple
access technique in wireless sensor networks,” IEEE Trans. Comput., vol. 60, no. 3, pp. 400–417,
2011.
[12] P. Toth and D. Vigo, “The Vehicle Routing Problem,” Soc. Ind. Appl. Math., 2001.
[13] M. M. Solomon, “ALGORITHMS FOR AND SCHEDULING PROBLEMS THE VEHICLE
ROUTING WITH TIME WINDOW CONSTRAINTSS,” Oper. Res., vol. 35, no. 2, pp. 254–265,
2010.
[14] B. Golden, L. Levy, and R. Vohra, “The orienteering problem,” Nav. Res. Logist., vol. 34, no. 3, pp.
307–318, 2006.

78
[15] M. Ma, Y. Yang, and M. Zhao, “Tour Planning for Mobile Data-Gathering Mechanisms in Wireless
Sensor Networks,” IEEE Trans. Veh. Technol., 2013.
[16] G. Xing, T. Wang, W. Jia, and M. Li, “Rendezvous design algorithms for wireless sensor networks
with a mobile base station,” in Proceedings of the 9th ACM international symposium on Mobile ad
hoc networking and computing (MobiHoc), 2008, pp. 231–240.
[17] G. Xing, T. Wang, Z. Xie, and W. Jia, “Rendezvous planning in mobility-assisted wireless sensor
networks,” in proceedings of the 28th IEEE InternationalReal-Time Systems Symposium (RTSS),
2007, pp. 311 – 320.
[18] N. Christofides, “Worst-case analysis of a new heuristic for the traveling salesman problem,” 1976.
[19] J. C. Hou, L. Kung, N. Li, H. Zhang, W. Chen, H. Tyan, and H. Lim, “J-Sim: A Simulation and
emulation environment for wireless sensor networks,” IEEE Wirel. Commun. Mag., vol. 13, no. 4, pp.
104–119, 2006.
[20] J. L. Hill and D. E. Culler, “Mica: A wireless platform for deeply embedded networks,” IEEE micro,
vol. 22, no. 6, pp. 12–24, 2002.
AUTHORS
Bassam A. Y. Alqaralleh received his BSc degree in Computer Science in 1992, graduate
Diploma in Computer Information Systemsin 2002 and Masters Degree in Computer
Science in 2004. After that, he received Ph. D. in Computer Science from University of
Sydney in 2010. His research interests are in the areas of Distributed Systems, Load-
Balancing, Networking and Security Systems. He has published a number of conference
and journal papers. Currently, he is an assistant professor at the Computer Science Department / Faculty of
Information Technology - Al-Hussein Bin Talal University. Since 2013, he is the dean for the Faculty of
Information Technology.
Khaled Almiani is currently an assistant professor at the school of computer Science, Al-
Hussein Bin Talal University, Jordan. His main research interests include designing
efficient algorithms to improve the performance of WSNs and game-theoretical
modellingfor WSNs. Khaled received his PhD degree in Information technology from the
University of Sydney in 2010.

Path constrained data gathering scheme for wireless sensor networks with mobile elements

More Related Content

What's hot (17)

Viewers also liked (19)

Similar to Path constrained data gathering scheme for wireless sensor networks with mobile elements (20)

Recently uploaded (20)

Path constrained data gathering scheme for wireless sensor networks with mobile elements