Proceedings of 2015 IEEE 12th International Conference on Networking, Sensing and Control Howard Civil Service International House, Taipei, Taiwan, April 9-11, 2015
Data Gathering in Wireless Sensor Networks with Ferry Nodes
1 1 Mariam Alnuaimi ,Khaled Shuaib \ Klaithem Alnuaimi and Mohammed Abed-Hafez l
2
College of Information Technology, UAEU P,O, Box 15551 Al Ain, UAE
e-mail: {mariam.aluuaimi.kshuaib.klaithem.nuaimi}@uaeu.ac.ae 2
College of Engineering, Electrical Engineering Department, UAEU P.O. Box 15551, AI Ain, UAE e-mail:
[email protected]
Abstract-
improve the performance such as the life time of a WSN and
Mobile ferries are an alternative way for
the maintained coverage area. Ferries are mobile elements
collecting data from dispersed sensor nodes especially in large scale
networks.
Unlike
data
collection
via
that are used to carry data over distance to the base stations
multi-hop
or to a data center.
forwarding among the nodes, ferries travel across the sensing
They are also used to connect isolated
islands of WSNs. In addition, ferries can be used to resolve
field and collect data from the sensing nodes. The advantage of
the issue of coverage for holes in a WSN resulting from the
using a ferry based approach is that, it eliminates the need for
need for replacing deployed fixed sensor nodes which have
multi-hop
energy
run out of energy. Mobile elements can be attached to
consumption at the nodes is significantly reduced. However,
people, animals, vehicles, robots, unmanned aerial vehicles
this increases data delivery latency and as such might not be
or any movable object. .
forwarding
of
data,
and
as
a
result
suitable for all applications. In this paper we survey the recent
There are different types of ferries or mobile elements
progress of using mobile ferry nodes for data gathering in
that are used in WSNs [7]. They can be classified as in the
WSNs addressing two main areas, determining the path of the
following sub sections:
ferry and the scheduling when to dispatch the ferry to collect data from sensors.
We also highlight challenges facing the
1.
deployment of mobile ferries in wireless sensor networks.
Ordinary sensor nodes: Ordinary sensor nodes are the source nodes that perform the sensing task as shown in Figure 1. Mobile ferry can be used as a scale sensor that
Keywords- ferry protocol; wireless sensor networks; load balancing; routing protocols; energy efficiency protocols;
senses data from the surrounding environment and sends the data (e.g, temperature, light, gas) to the cluster head or a collector. The advantage of these nodes is that they are moving, so they can track a
I. INTRODUCTION
movable event like intruder detection in border
Recent improvement in hardware electronics technology
monitoring applications.
enabled manufacturers to develop low cost, low power and small size sensors [1, 2, 3]. Hundreds and thousands of these sensors are deployed as wireless sensor networks (WSN) serving many applications based on the specific requirements of each one [4, 5, 6]. Nowadays, there are many applications of
sensor
networks
agriculture, toys,
covering
medicine,
intrusion
military,
detection,
different
fields
environment
motion
such
as
monitoring,
tracking,
machine
malfunction, and many others [2]. Sensors can be deployed to continuously report environmental data for long periods of time. In general, a wireless sensor network is a collection of static
nodes
with
sensing,
computation,
and
wireless
Base station
communication capabilities [2]. However, due to the nature of some applications, mobile nodes are needed. Using
Figure 1 Mobile node is used as a part ofWSNs
mobile nodes to collect data from sensors in WSNs can
978-1-4799-8069-7/15/$3\.00 ©20151EEE
221
II.
Mobile Sink or Base Station
2.
ApPLICATIONS OF USING MOBILE FERRIES INWSNs
The Sink or the Base Station (BS): destination where
Due to the nature of some applications of WSNs, mobile
all data are gathered to be used by data centers or
ferries are needed. Using ferries to collect data from sensors
outside applications. Sink can be mobile and visit all
in WSNs can improve the performance of WSNs such as the
nodes to collect data from them directly or through
life
intermediate nodes as shown in Figure 2. Mobile
applications that can utilize the advantages of using ferries.
sinks can increase the network lifetime, decrease delay
and
decrease
traffic.
However,
having
time
1.
a
and
the
coverage.
Below
are
some
of
the
Border monitoring
Mobile node can be used in intrusion detection and
mobile sink requires a full knowledge or control
border surveillance to collect data from sensors scattered
over its movement and schedule.
along the border line. BorderSense is an example of such an application, where mobile ferry such as Unmanned Aerial Vehicles (UAVs) are used to collect data from static sensors
Mobile Sink
[19]. Mobile nodes can also be used as sensor nodes to provide additional coverage in case needed. In addition to that, mobile nodes can track intruders based on information from static sensors and help in catching the intruders. 2.
Disaster Recovery
During times of a disaster, such as an earthquake or a tsunami
the
communication
infrastructures
usually
are
destroyed, which made rescue and recovery efforts hard. Therefore, there is a need for mobile nodes to be used in collecting information from the surrounding environment and help in rescuing. In [20] multiple mobile sensors carried on vehicles are used across large distances with minimal need for wired infrastructures to provide communication coverage for disaster recovery. In addition, static nodes can Figure 2 Mobile Sink collects data from nodes
be deployed to monitor the disaster area and information can either be disseminated through multi-hop forwarding or by
Mobile support nodes
3.
using mobile ferries carried by robots or other means.
Support nodes: the intennediate nodes that help the data to
be
transferred
from
the
source
(sensing
nodes)
3.
to
Environment Monitoring
destination (Sink) as shown in Figure 3. A WSN might
Mobile elements can be attached to people, animals,
become partitioned into several islands for many reasons,
vehicles or any movable object to continuously report
which makes the communication in the network impossible.
environmental data for long periods of time. They can be
In this case, mobile support nodes can be used to connect
used in
portioned WSN. Mobile support node can also be used to
detection flood detection. CitiSense
detecting
air and water
pollutions,
forest
fire
[21] uses wearable devices and mobile phones carried by users to collect
replace dead critical nodes in case of emergency. An example of it is sending a robot to cover a certain area when
environmental parameters (CO, N02 and 03, temperature,
nodes in duty are dead. This strategy will help in providing
humidity and barometric pressure) from static sensors to
more coverage and increase the lifetime of the network.
monitor air pollution in certain areas and associate them with other events or aspects. 4.
�
NOde
��
.•
Node
battlefield Nod�._.....-�
�.
•.
No!
•..
..
\
.d
Military Applications
Military
applications surveillance,
involve
intrusion
monitoring
friendly
detection, forces,
battlefield damage assessments, information gathering and
�'V"'�'
smart logistics support in an unknown deployed area. In [22] static sensor nodes deployed on the ground with a job to detect and track vehicles passing through the area over a dirt
�
road. The vehicle tracking information was collected from
NO
the sensors using the UAV in a flyover maneuver and then sent the data to an observer at the base camp.
Base station
5.
Intelligent Road Transportation
Applications of Intelligent Road Transportation (lRT)
Figure 3 Mobile support node used to transfer data in
usually fall under navigation,
WSNs
222
traffic flow control
(e.g.
changing traffic lights) and the need to plan and build new
collect sensed data whenever they are in the communication
infrastructures.
range of the static sensor nodes. In [8] mobile entities called mules were deployed in the environment. Mules picked up
Vehicles on the roads equipped with sensors could act as
data from sensors when they were in close range, buffered it,
mobile nodes on the road network providing a rich source of
and dropped it off when within a communication range of
data about the traffic, environment and road conditions. This
wired access points. They used two-dimensional random
information
traffic
walk to model the mobility of mules. Both mules and sensors
effectively in order to maintain a good traffic flow and
required to have memory capacities as they were buffering
minimize the risk of accidents and roads congestions [26]. In addition, data can be disseminated by these mobile nodes to
data. In [9] mobile nodes were used in the sensor field as forwarding agents. When a mobile node moved in close
assists
traffic
managers
to
regulate
subscribed drivers to avoid congestion and get road and
proximity to sensors, data is transferred to the mobile node
weather conditions in real-time at a minimum cost.
for later depositing at the destination. They used analytical models to understand the key performance metrics such as
6.
Animal Tracking
data
Sensors can be attached to animals to track them in
transfer,
latency
to
the
destination,
and
power
consumption.
support of wildlife research or simply to locate them. When
Due to the random mobility of the ferry it is difficult to
sensors are attached to animals they became mobile sensor
gather sensed data from all deployed nodes. Unlike the
nodes. ZebraNet system is a WSNs tracking system carried
random path approach, in the planned path approach a path is
by animals under study across a large area. Sensors carry
determined before the dispatching of the ferry and thus the
logged data to researchers such as the animal positions, their
ferry is sent to cover a certain area of deployed sensors to
temperature, heart rate and frequency of feeding and send
collect data. In [10], an architecture of a wireless sensor network for a traffic surveillance application with mobile
them to the base center to be used by wildlife researchers
sinks was proposed. All sensor nodes in this architecture
[27]. These mobile sensors can also be used to collect data from scattered static sensors deployed in the farm area for
were assumed to be located within the direct communication
various applications and communicate such data effectively
range of the mobile sink. All multi-hop transmissions of high
to a base station for further transmission to a central database
volume data over the network are converted into single-hop
and processing.
transmissions to preserve the energy of the network further. Therefore; nodes will transmit only in a single-hop fashion to
7.
the mobile sink.
Pipeline Monitoring
There
are
many
applications
of WSNs
in
pipeline
monitoring
and
oil
nodes
used
in
In Mobi-Route [11], a routing protocol where the sink moves on a planned path to prolong the network lifetime in
pipelines monitoring because pipelines are stretched along a
WSNs was proposed. In this protocol, the sink moved and
large area and it is more costly to deploy static nodes across
stopped at certain points of interest. The stopping time
monitoring pipeline
such
as
monitoring.
water
pipeline
Mobile
sensor
are
it. TriopusNet is a mobile wireless sensor network system for
periods were designed to be long enough to allow for the
autonomous sensor deployment in pipeline monitoring. It
collection of data. All deployed static sensor nodes needed to
releases sensor nodes from a centralized repository located at
be aware of the sink movement and the location and time of
the source of the water pipeline building a wireless network
stops to send the sensed data to it.
of interconnected sensor nodes. When a node dies or has a
The authors in [12] used a single ferry to collect data from a circular dense sensors network. They showed that the
low battery level TriopusNet system sends a new node from
optimal mobility strategy of the ferry achieved when moving
the repository to replace the dead node [28].
at the border of the sensing area. They divided the area into III.
circles starting from the source. The inner circles forward the SURVEYING PREVIOUS RESEARCH WORK
data to the outer ones; until the border was reached where the ferry was used to collect the sensed data.
Using mobile ferry in WSNs is relatively a new area of research which is gaining the attention of many researchers.
2.
Incorporating ferries in WSNs helps in eliminating the need
The scheduling of when exactly to send the ferry to
for multi-hop forwarding of data. It also reduces the energy
collect sensed data from nodes is a quiet complicated task. In
consumption at the nodes. However, using ferries might add
[13, 14] authors studied the scheduling problem where the path of the mobile sink was optimized to visit each node in
delay in collecting, disseminating and processing data and thus might not be suitable for all applications. Existing
the WSN before its buffer was full. Buffer overflow was
research in this field can be classified in two main areas or
used as a trigger to send the ferry to collect data to avoid data
categories as listed in the below subsections. 1.
Scheduling the dispatch of the ferry
loss.
Determining the path
In [15] the authors suggested the mobile sink to visit locations (rendezvous points) based-on a
exact
The path that the ferry takes to collect sensed data from
predetermined schedule to collect data. The rendezvous
sensors can be classified into two categories, random path
points buffer and aggregate data originated from the source
and planned path. Usually in case of the random path the ferry is attached to people or animals moving randomly to
223
nodes through multi-hopping and transfer it to the mobile
2.
sink upon its arrival. In[16] a ferry is used to help in collecting data in partitioned
wireless
sensor
networks
and
transfer
carry data over distances to the base stations. Their mobility
the
can be an issue when collecting the data from sensors.
collected data stored locally back to the base station. Authors
Therefore, presence detection, speed and direction of the
classified scheduling of ferry visit into three categories, time
mobile ferry to enable nodes to send data and the ferry to
based scheduling, location based scheduling and dynamic
collect data from the nodes in an efficient way while
based scheduling. Time based scheduling happened when a
preserving as much energy of the network nodes as possible
node dies and its death leads to portioned WSNs. This node
is an important challenge that needs to be investigated.
will have a higher priority for ferry visits. The location based
3.
scheduling assigns the nodes closer to the base station a higher priority for the ferry visit. The dynamic based
visited by the ferry yet, and selects the shortest distance for
In this research, sensor nodes broadcast data collection
have to be defmed to improve the performance of the network. This will allow nodes to sleep and awake based on
In [18] a mobile node was used to help in disseminating data to the sink. It was used to move back and forth along the
the ferry schedule and proximity. As a result, energy of the network will be preserved and the life time of the network
linear network, and collect data from the individual sensors
will increase. If the path and schedule of the ferry is known
when it comes within their communication range. The
or can be predicted, sensors can be awakened only when they
mobile node will then transfer the collected data to a base
expect the ferry to be in their communication range. This
station. The mobile node was also used to perform other
will further preserve the energy of the network and it is still
functions, such as data processing, and aggregation, and can
an open research challenge to optimize the motion of the
also transport messages from the sink to the sensor nodes.
ferry in a controlled manner and thus efficiently manage the
Table 1 shows a summary literature work of using ferries
duty cycles of sensing nodes depending on the deployed
in WSNs. The research work is classified into either using
application.
single-hop or multi-hop in forwarding data. Furthermore, the
4.
research is classified into two categories either all nodes in the network are visited by the ferry or only a subset of nodes.
[15], [17], [11], [12]
IV.
Optimum data transfer
The communication time between static nodes and the mobile ferry might be short while a significant amount of
re ate d work
[9]
data might need to be collected. Therefore, there is a need to provide coverage to the entire network and maximize the number of reliable sent massages to the ferry. In [25] authors investigated how to efficiently collect data from stationary sensor nodes using multiple robotic vehicles such as data ferries under different circumstances. They proved that
CHALLENGES IN USING FERRY IN WSNS
finding an optimum ferry path is an np-hard problem. Therefore, finding a reliable and efficient path for the ferry
Based on the previous background and literature review
to take to provide coverage for the entire network using least
discussed several challenges can be identified in deploying
energy and minimum latency is one of the hardest challenges
ferries inWSNs. Below are some of these challenges that are
that needs to be investigated and addressed.
still open research areas to be tackled in the future by researchers. 1.
all
when the motion of the ferry is controlled. As an example,
collect data, and transfer the data to the sink.
Multi-hop
in
the path, the speed and the stopping time periods of the ferry
such requests, the ferry moves toward the sensor nodes to
[16], [8]
issue
mobile ferry to visit sensors in a WSN should be defined
requests when their buffers are about to be full. On receiving
[13], [14], [18], [10]
important
the awake mode consumes energy. A predefmed policy for a
In [17] the authors considered on demand data collection.
single-hop
an
static and mobile nodes. They found that keeping nodes in
next visit.
Visits subset of nodes
is
network life time ofWSNs. Authors in [23, 24] surveyed the existing energy management schemes in literature for both
partitioned wireless sensor networks that have not been
0f
management
networks. Efficient management can lead to prolonging the
current location of the ferry and the locations of the
Visits whole nodes
Ef f icient Energy Management Strategies
Energy
scheduling is based on calculating the distances between the
Table 1 Summary
Mobility of the ferry
Ferries are rechargeable mobile elements that are used to
V.
Ferry Presence detection
CONCLUSIONS
In this paper we have surveyed the recent progress of
Detecting the presence of the ferry in the communication
using mobile ferries for data gathering in WSNs addressing
range is a very challenging issue especially if the presence
two areas,
was for a very short time and the path of the ferry is
scheduling when to dispatch the ferry to collect data from
uncontrollable. Therefore, sensors need to be awake and in
static sensors. We presented a classification of mobile
detection mode all the time to detect the presence of the
ferries, based on the role they play in addition to carrying
ferry.
determining
the path of the ferry and the
information. Furthermore, we surveyed the existing work on path planning and scheduling of ferry dispatching in the literature.
224
In
addition
to
that,
common
challenges
in
[16] Chen, Tzung-Cheng, Tzung-Shi Chen, and Ping-Wen Wu.
deploying mobile ferries inWSNs were discussed along with
"Data collection in wireless sensor networks assisted by mobile collector." Wireless Days, 2008. WD'08. 1st IFIP. IEEE, 2008.
many of their possible applications.
[17] He, Liang, et al. "Evaluating on-demand data collection with
REFERENCES
mobile elements in wireless sensor networks." Vehicular technology conference Fall (VTC 201O-Fall), 2010 IEEE 72nd. IEEE, 2010.
[I]
I.Akyildiz, W.Su , Y.Sankarasubramaniam, E. Cayirci, "Wireless sensor networks: a survey", Computer networks 38, no. 4 (2002): 393-422. Jan,2002.
[2]
[18] Imad Jawhar, Mostafa Ammar, Sheng Zhang, Jie Wu, Nader
P.Baronti, P. Pillai, V.Chook"V. Chessa, AGotta, "Wireless sensor networks: a survey on the state of the art and the 802.15.4 and ZigBee standards". Computer communications 30, no. 7 (2007): 1655-1695, 2007.
Mohamed, "Ferry-Based Linear WirelessWireless Sensor Networks", accepted in the IEEE Global Communications Conference Exhibition & Industry Forum (IEEE Globecom 2013), Atlanta, Georgia, USA, December 9-13, 2013.
[3]
M. Alnuaimi, F. Sallabi, and K. Shuaib, "A Survey of Wireless Multimedia Sensor Networks Challenges and Solutions", IEEE International Conference on Innovations in Information Technology (lIT' 11), pp. 191-196, Abu Dhabi, UAE, April 2011.
[19] Z. Sun, P. Wang, M.C. Vuran, M. AI-Rodhaan, A AI
[4]
M. Alnuaimi, K. Shuaib, I. Jawhar, "Performance Evaluation of IEEE 802.15.4 Physical Layer Using MatLab/Simulink", IEEE International Conference on Innovations in Information Technology, 2006 , pp.I-5, Nov. 19-21, 2006.
"Communications support for disaster recovery operations using hybrid mobile ad-hoc networks," in Proceedings of the 32nd IEEE Conference on Local Computer Networks, 2007.
[5]
[21] Nikzad, N., Verma, N., Ziftci, C., Bales, E., Quick, N.'vZappi,
K. Alnuaimi, M. Alnuaimi, N. Mohamed, I. Jawhar, K. Shuaib, "Web-based wireless sensor networks: a survey of architectures and applications. The 6th International Conference on Ubiquitous Information Management and Communication, Feb. 20-22, Malaysia, 2012.
P., Patrick, K., Dasgupta, S.,5Krueger, I., Rosing, T.S., and Griswold, W. G.: "CitiSense: Improving Geospatial Environmental Assessment of Air Quality Using a Wireless Personal Exposure Monitoring System", Wireless Health, 2012.
[6]
Garcia Villalba, Luis Javier, et al. "Routing protocols in wireless sensor networks" Sensors 9, no. 11 (2009): 83998421, Oct 2009.
[7]
Di Francesco, Mario, Sajal K. Das, and Giuseppe Anastasi. "Data collection in wireless sensor networks with mobile elements: A survey." ACM Transactions on Sensor Networks (TOSN) 8.1, 20II.
[8]
[9]
Dhelaan, I.F. Akyildiz, BorderSense: border patrol through advanced wireless sensor networks, Ad Hoc Networks 9, 2011.
[20] W. Lu, W. K. G. Seah, E. W. C. Peh, and Y. Ge,
[22] Arora,
Anish, Prabal Dutta, Sandip Bapat, Vinod Kulathumani, Hongwei Zhang, Vinayak Naik, Vineet Mittal et al. "A line in the sand: a wireless sensor network for target detection, classification, and tracking." Computer Networks 46, no. 5, 2004.
[23] G. Anastasi, M. Conti, M. Di Francesco, A Passarella, "Energy Conservation in Wireless Sensor Networks", Ad Hoc Networks, to appear(available at http://info.iet.unipi.it/�anastasi/papers/adhoc08.pdf).
R.C. Shah, S. Roy, S. Jain, and W. Brunette, "Data MULEs: Modeling a Three-Tier Architecture for Sparse Sensor Networks,"Proc. First IEEE Int'l Workshop Sensor Network Protocols and Applications, pp. 30-41, 2003.
[24] V. Raghunathan, S. Ganeriwal, M. Srivastava, "Emerging Techniques for Long Lived Wireless Sensor Networks", IEEE Communications Magazine, Vol. 44, N. 4, April 2006.
S. Jain, R.C. Shah, W. Brunette, G. Borriello, and S. Roy,"Exploiting Mobility for Energy Efficient Data Collection in Sensor Networks," Mobile Networks and Applications, vol. 11, no. 3, pp. 327-339, 2006.
[25] Xue, Lirong, Donghyun Kim, Yuqing Zhu, Deying Li, Wei Wang, and Alade O. Tokuta. "MUltiple Heterogeneous Data Ferry Trajectory Planning in Wireless Sensor Networks." , INFOCOM, May, 2014.
[10] L. Song and D. Hatzinakos, "Architecture of Wireless Sensor
[26] Z. Koppanyi, T. Lovas, A Barsi, H. Demeter, A Beeharee,
Networks with Mobile Sinks: Sparsely Deployed Sensors," IEEE Trans. Vehicular Technology, vol. 56, no. 4, pp. 18261836, July 2007.
and A Berenyi. "Tracking vehicle in gsm network to support intelligent transportation systems." Int. Arch. Photogramm. Remote Sens. Spatial Inf. Sci., XXXIX-B2:139-144, 2012. ISPRS Archives.
[II] J. Luo, J. Panchard, M. Piorkowski, M. Grossglauser, and J. Hubaux, "MobiRoute: Routing towards a Mobile Sink for Improving Lifetime in Sensor Networks," Proc. Second IEEEIACM Int'l Conf. Distributed Computing in Sensor Systems (DCOSS), pp. 480-497, 2006.
[27] P. Juang, H. Oki, Y. Wang, M. Maronosi, L. Peh, D. Rubenstein, "Energy-Efficient Computing for Wildlife Tracking: Design Tradeoffs and Early Experiences with ZebraNet", Proc. ASPLOS, Oct 2002.
[12] Luo, J. and Hubaux, J.-P. 2005. "Joint mobility and routing
[28] T. T.-T. Lai, W.-J. Chen, K.-H. Li, P. Huang, and H.H. Chu.
for lifetime elongation in wireless sensor networks". Proceedings of the 24th IEEE Conference on Computer Communications (INFOCOM 2005). Vol. 3. 1735, 2005
Triopusnet: automating wireless sensor network deployment and replacement in pipeline monitoring. In Proc. of ACM/IEEE IPSN. ACM, 2012
[13] A Somasundara, A Ramamoorthy, and M. Srivastava, "Mobile Element Scheduling for Efficient Data Collection in Wireless Sensor Networks with Dynamic Deadlines," Proc. 25th IEEE Int'l Real-Time Systems Symp. (RTSS), pp. 296305, 2004.
[14] Y. Gu, D. Bozdag, R. Brewer, and E. Ekici, "Data Harvesting with Mobile Elements in Wireless Sensor Networks," Computer Networks, vol. 50, no. 17, pp. 3449-3465, 2006.
[15] G. Xing, T. Wang, Z. Xie, and W. Jia, "Rendezvous Design Algorithms for Wireless Sensor Networks with a Mobile Base Station," Proc. ACM MobiHoc, pp. 231-240, 2008.
225