International Journal of Grid and Distributed Computing

Vol. 5, No. 4, December, 2012

An Context-aware Management and Control Mechanism in a Mobile

Ad-Hoc Environment

Qi Liu1, Nigel Linge2, Jin Wang1, Sai Ji1 and Ting Huai Ma1*
1
Jiangsu Engineering Center of Network Monitoring, Nanjing University of
Information Science & Technology, Nanjing, 210044, China
2
Computer Networking and Telecommunications Research Centre, University of
Salford, Salford, Greater Manchester, M5 4WT, UK
[email protected]

Abstract
Complete integration and interaction of network objects can be achieved in a mobile ad-
hoc environment. Yet, the dynamic nature of free movement and interaction among these
objects also brings challenges on information representation, processing and communication.
This paper presents a context-aware mechanism, which is able to implement efficient
information exchange and management between nodes in a mobile ad-hoc network (MANET).
Rejoining and information loop issues happened heavily in a mobile ad-hoc environment has
been discussed in this paper. According to the simulation results, our approach to context-
aware management and control makes the information loop problem well controlled.

Keywords: Information Loop Control, Context Awareness, Rejoining Control, Mobile Ad-
Hoc Networks

1. Introduction
Recent research on context-aware middleware brings rapid growth in the number of mobile
applications, where a device’s context information has been explored including its location,
type, technical capability, functions, nearby devices and more general information about the
physical environment [1]. For ad-hoc networks there too is a requirement to be able to detect,
share and respond to contextual information. However, the unpredictability and dynamically
changing topology, especially when one introduces mobility, present particular challenges for
the reliable and timely dissemination of context information.
This paper introduces a new method to manage and control context exchange within an ad-
hoc network. The framework comprises two components, a context information database that
provides context representation and storage, and a communication protocol called CiComm
used for exchanging context information. In such environment, efficient management
strategies are established in order to eliminates potential problems arisen by random node
movement, e.g. information loops, rejoining, etc.
The remainder of this paper is organized as follows. Related work on context awareness
and its application for ad-hoc networks is presented in Section 2. Section 3 introduces our
framework with its features and functions. Simulation results are discussed in Section 4 and
Section 5 concludes the paper and evaluates the framework.

21
International Journal of Grid and Distributed Computing
Vol. 5, No. 4, December, 2012

2. Related Work
The concept of context awareness to a mobile ad hoc network is implemented and
applied using two main categories. Some prefer context information as integrated
components to provide internal application program interfaces for the invocation of
upper and lower layers. In these applications, environmental information and relevant
properties are regarded and recorded as context information. Alternatively, complete
frameworks have been also considered and designed to provide a comprehensive
context awareness based solution. In this section, the applications of context aware are
firstly reviewed, followed by both categories mentioned above. After that, service
discovery protocols are also discussed compared with information discovery in context -
aware applications.

2.1. Context-Aware Applications

Context awareness can be applied into most domains of computer science and networking.
For example, a context-aware routing protocol called Communication Inter Vehicle
Intelligent Cooperative (CIVIC) was designed for the communication between vehicles [2].
The CIVIC was implemented in LiveNode sensors with LIMOS [3]. Eigner and Mair reported
their work in [4], where a Vehicular Ad Hoc Network (VANET) was defined for ad hoc
communication between vehicles at high speed.
Vertical soft handover for better performance was discussed in [5], which used context-
aware parameters to achieve the goals. The context parameters in the work included user
required bandwidth, user traffic cost, access network utilization, and user received Signal to
Interface-plus-Noise Ratio (SINR). In [6], context information was used to represent user
information in a mobile network including their behaviors, mobility, network usage, device
type, etc.. This information was then converted into Keyhole Markup Language (KML)
scripts and virtualized in the Google Earth software.

2.2. Context-aware Components

A de-facto definition of context and context awareness was given by Dey [7, 8],
which identifies all relevant information that is able to describe the “situation of an
entity” as its context. This definition has been well used in context-awware components,
which identified context as explicit parameters, e.g. devices, users, locations, time, the
network. Semantic languages such as XML, OWL, etc. were used to represent context
information in these designs [9].
In [10], a cross-layer routing protocol was presented in MANETs, where context
information is defined to include energy consumption for better route discovery, such as
TX and RX power consumption, Angle of Arrival (AoA), etc. Better results were
concluded for more context information had been collected and processed.
A schema calculating a trust value of nodes in MANET was designed in [11]. The
considered parameters included previous interactions, observation of present behavior,
recommendation from direct paired devices, etc. The collection can be recognized as
context information to build general nodes trust.
A location-aware algorithm was presented to broadcast messages in a mobile ad hoc
network [12]. Different zones were defined based on neighbor positions to allow or
deny forwarding. The results indicate better performance than other broadcasting
algorithms in low densities.
Other research work involves a self-policing reputation mechanism that values
nodes’ neighbors [13], a context-aware adaptive routing protocol (CAR) for best

22
 International Journal of Grid and Distributed Computing
Vol. 5, No. 4, December, 2012

forwarding route considering a node’s battery level, mobility and co-location etc. [14,
15], a context sensitive binding mechanism for better service migration [1 6], and a
decision support engine based on a general Bayesian network approach [17].

2.3. Context-aware Frameworks

Context-aware architectures were designed to provide complete representation and
sharing mechanism of context information. Compared to context-aware components
above, the solutions below provide comprehensive context-aware strategies/approaches
in a mobile ad hoc network.
Context-aware migratory services were offered based on a Smart Messages platform
in a ubiquitous environment [18], where Ccontext information was interpreted
(MonitoredCxt), stored and shared (Context Manager), and evaluated (Validator).
A CoBrA system (Context Broker Architecture) provided an agent based architecture
featuring contextual information in pervasive spaces. A context model based on a
context-aware message broker was designed to offer common interfaces to context-
aware applications [19]. In this system, a context-aware conceptual model federated
multiple context brokers was designed to facilitate context provision and usage
efficiently [20].
In [21], a platform called ACORD-CS that supports full deployment of context-aware
applications was presented, including a meta-data model identifying potential context
information, and a middleware providing APIs to context-aware applications.
Other models and solutions include a Reference Model for a MANET environment
[22], the LMSS system to provide relevant location information of mobile users [23],
and a context-aware architecture for Service-Oriented Computing [24].

2.4. Context Awareness and Service Discovery Architectures

Service discovery protocols (SDPs) were proposed for the purposes of the automatic
discovery of the presence of services, the determination of where and who to provide
the services and the management of the retrieval and execution of these services. In
terms of network services, Sun’s Jini technology explains it as “an entity that can be
used by a person, a program, or another service”. Two examples were given as services,
including a printing job and a translation process. In our paper, the exchange of context
information is regarded as a service of an ad-hoc network; and hence, an SDP protocol
can be used for context discovery and exchange.
In [25, 26] SDPs have independently proposed that they should be categorized as
directory based and directory-less. Directory servers are used in Directory based SDPs
to maintain the information of devices and their services. These SDPs can be further
classified as either distributed or centralized based on their database types. Directory-
less SDPs discover services by sending broadcast or multicast requests to neighboring
devices. In general a directory based SDP, e.g. Jini and DSDP [27] fits large scale
networks with low levels of device mobility whilst directory-less SDPs are more suited
to small scale networks or networks with high levels of device mobility. Another
example is that of the Service Location Protocol (SLP) which has been adapted for use
in mobile ad-hoc networks, resulting in the creation of SLPManet [28, 29].

3. Strategies to Information Management in MANET

In mobile ad-hoc networks devices can exchange context if their behavior and attributes
can be justified and decided upon by locally retrieved and remotely shared context

23
International Journal of Grid and Distributed Computing
Vol. 5, No. 4, December, 2012

information. To enable the exchange of such contextual information we propose a method

comprising three parts: a representation model, a context information repository and a
communication protocol. The model defines context information in an ad-hoc network
environment. It introduces a scheme for context representation and establishes a hierarchical
structure to categorize context parameters. The predefined context parameters can then be
maintained and shared via the repository and communication protocol.

3.1. Context-aware Model

A complete representation of relevant information requires a generic method with common
rules for abstraction of the information. To meet this requirement, an ontological solution is
presented in this thesis. The ontology defines five basic interrogative dimensions; they are
Who, When, Where, What and How.
Each dimension retrieves one aspect of relevant information of mobile devices within an
ad hoc network. The “who, when, where, what and how” questions are raised to recognize
and categorize potential information within an ad hoc network. These devices’ information
can correspondingly be abstracted into parameters, separated in each dimension. This
structure allows for context parameters from other devices within the ad hoc network to be
added without interference; thereby providing more flexibility in maintaining these
parameters.

3.2. Context Information Base and Communication

A context-aware framework previously designed for the representation, storage and
communication of contextual information [30] has been employed in this paper to exchange
relevant information between mobile devices in a mobile ad-hoc network. Firstly, a
hierarchical structure is maintained underneath each interrogative dimension. In addition, an
index scheme is conducted to assign a set of reference numbers to corresponding parameters
according to their position at the structure. Following the design of the context-aware
framework, the predefined parameters are stored in an information repository called context
information base (CiB). These parameters can then be shared among mobile device in a
MANET via the CiComm protocol.
Relevant information can be exchanged between mobile devices in a MANET. The
management of information communication can be separated into three steps including
Preparation, Update and Upgrade. In the Preparation step, the mobile device will inspect the
integrity of the local CiB structure, collect local context information for further exchange, and
then initialize the neighbor list. After this, a SHARE packet is sent to new known neighbors.
In the Update step, the CiData Table of the local CiB will be updated including assigning a
neighbor index number, and then creating/updating neighbor entries in the CiData table. In
the Upgrade step, the device will be registered into the receiver, which is achieved by
upgrading the local CiSchema table. Two entries are created containing the neighbor’s
identity and the index number of the retrieved context parameter. In the case of entries having
been created, the existing entries will be amended by updating the list of index numbers.

3.3. Information Management and Loop Control

A lease control is introduced to help maintain the relevant information from proximity and
kept it up to date, as shown in Figure 1. By default, the lease timer is set as five times long as
the heart-beating timer. Failure to receive periodically incoming heart-beating beacons for
more than five times will cause the removal of the neighbor and the associated entries
including all direct context from that neighbor in both CiData and CiSchema tables. The lease

24
 International Journal of Grid and Distributed Computing
Vol. 5, No. 4, December, 2012

timeout value is deduced by one every time when the heart-beat occurs; whereas the value is
reset/renown when a new Beacon_HB frame is received from the corresponding mobile
device. The mechanism therefore actively controls neighbors’ movements i.e. their departure
and rejoining.

Figure 1. Lease Control within a Local Repository. When Node B moves out of
the communication range of Node A, its beacon cannot reach Node A, and
hence its relevant context information is removed from Node A

4. Simulation and Results

The context-aware framework has been implemented in NS2. Particular scenarios have
been designed to evaluate the impact of loops and dead loops, and rejoining.

4.1. Scenario One: Copping with Information Loops and Dead Loops
In this scenario, two simulations are designed to examine the potential loop back and dead
loop issues. Here a loop means same information is shared back to the source device via one
or more intermediate devices. A dead loop means two or more devices continuously send the
same request and response packets to each other. The scenario settings are shown in Table 1,
whereas the topology is shown in Figure 2.

Table 1. Simulation Settings for Scenario One

Simulation Settings Simulation 1 Simulation 2
2
Simulation Area 10x10 m 20x20 m2
Simulation Duration 30 seconds 30 seconds
Number of Devices 3 4
Bandwidth 11Mb 11Mb
Communication Range 10 meters 10 meters
Heart Beat Timer 1 second 1 second
Lease Timer 2 seconds 2 seconds
CiB Load and Release time random random

25
International Journal of Grid and Distributed Computing
Vol. 5, No. 4, December, 2012

(a) Context information looping back via an intermediate node

(b) Context information continuous updating at Node 3

Figure 2. Topologies of Two Simulations. The upper topology (a) depicts how
information of device 0 is looped back to itself. The bottom one (b)
demonstrates information of device 0 is continuously updated at device 3.

The list of total UPGRADE packets generated in Simulation 1 is captured and shown in
Figure 3. Four columns are created in the list; they are, from left to right, the receiving time,
the source and destination CiB address, the source and destination MAC address, and the
packet body.

Figure 3. Results of Simulation 1. Six UPGRADE packets were captured.

According to the fact that in each packet, the next-hop address tagged in the
MAC Address field differs from the destination address recorded in the packet
body, it can concluded that no looping back happened during the simulation

26
 International Journal of Grid and Distributed Computing
Vol. 5, No. 4, December, 2012

The reason that only UPGRADE packets are displayed is because context information is
only looped back via this type of CiComm packets. Through the list, it can be seen that six
UPGRADE packets were generated. The device identity shown within the curly brackets in
each packet body is different from the corresponding destination address in the MAC Address
field, which proves that no packets are looped back to original devices. The CiComm protocol
in each device always checks the destination that it is sending a packet to. An UPGRADE
packet is only transmitted to those devices that are not the same as those drawn within the
curly brackets of the packet body.
Figure 4 summarizes the REQUEST packets generated in Simulation 2. According to
Figure 2(b), dead loops in this scenario happen only if context information of an intermediate
device is requested and there are two active routes to that device. In this figure, the four
entries shown in red are REQUEST packets that acquire context information from indirect
devices. This can be seen through the fact that their CiB destination addresses are different
from their MAC destination addresses. The limited number of REQUEST packets generated
for indirect context information proves that no dead loops happen within the simulation. In
this scenario, there are two active routes between device 0 and 3.
It is also noticed in Figure 4 that before requesting context information from Device 3 and
Device 0, Device 2 requested their information from other devices. In the figure, Device 2
sent a REQUEST packet to Device 1 containing Device 3’s context information “3.1.1” at
1.5318 seconds. This happened again at 1.5403 seconds but containing Device 0’s
information via Device 3 at that time. After these events, Device 2 then sent REQUEST
packets directly to Device 3 at 1.5395 seconds and Device 0 at 1.5952 seconds respectively.
Such situations happened due to actual reception order of Device 2. That is, before the
reception of context information directly from the original device (in this case, Device 3 at
1.5318 seconds), the destination received forwarded packets via other indirect device(s) (i.e.
Device 1).

Figure 4. Results of Simulation 2. REQUEST packets were captured. No dead

loop or looping back issues happened.

27
International Journal of Grid and Distributed Computing
Vol. 5, No. 4, December, 2012

Figure 5 shows SHARE packets transmitted and received by Device 2 in Simulation 2.

Through the figure, it can be seen that when Device 1 shared its context information with
Device 2 at 1.5279 seconds, context information of Device 0 and Device 3 was also contained,
which proves that Device 1 had retrieved context information from those devices. Meanwhile,
due to no SHARE packets about these two devices being received by Device 2 yet,
REQUEST packets were sent to Device 1 for these devices’ Information at 1.5318 and 1.5403
seconds respectively, (see in Figure 4). After that, context was received from directly
connected devices to Node 2 at 1.5384 and 1.5943 seconds, as shown in Figure 5.
Corresponding REQUEST packets were therefore generated again at 1.5395 and 1.5952
seconds, as shown in Figure 4.

Figure 5. Results of Simulation 2. SHARE packets were captured. No dead loop

or looping back issues happened.

Through Figure 4 and 5, it can be seen that late reception of context information from
directly connected devices can generate redundant CiComm packets, but this traffic does not
result in dead loops. Having received SHARE packets from Device 3 (at 1.5384 seconds) and
Device 0 (at 1.5943 seconds), Device 2 updated its local CiB with direct context information
and stopped requesting context from the intermediate device, i.e. Device 1. In addition,
Device 0 and 3 requested context from each other via Device 1 and no dead loop was created.

4.2. Scenario Two: Copping with Departure and Rejoining

This scenario aims to examine the rejoining behavior of devices in context driven ad hoc
networks. Here rejoining is defined as the process by which a device’s movement causes it to
move out of communication range with its neighbors and subsequently travelling back into
range, as shown in Figure 6.

28
 International Journal of Grid and Distributed Computing
Vol. 5, No. 4, December, 2012

Figure 6. Topologies of Two Simulations. The movement of Device 1 is

configured in three phases. The first one starts at 2.0 seconds, when Device 1
moves from position (50.0, 7.0) to (0.0, 7.0) at 8 meters per second. At 7.5
seconds, it changes its direction to its original position (50.0, 7.0) at the same
speed. Finally at 10.0 seconds, Device 1 moves back to position (0.0, 7.0) again
at 8 m/s and stops there until the simulation is over.

According to the movement of Device 1, it enters and leaves communication range with
Device 0 three times, as shown in Table 2. Because the speed of Device 1 is constant (8
metres per second), the duration that it stays in the communication range of Device 0 is
always 1.5 seconds.

Table 2. Entering and Leaving Device 0‟s Communication Range

Joining Time Leaving Time
1 5.625 7.125
2 7.875 9.375
3 10.625 12.125

In this scenario, three different lease timers are configured to demonstrate how the lease
timer controls potential redundant contextual traffic generated because of frequent rejoining
events. In Figure 7, SHARE packets transmitted between two devices are listed. These
packets can be separated into three groups according to the time stamps; these are transmitted
at about 6.5, 8.5 and 11.5 seconds respectively. The lease timer was set to 1 second, so every
time when the devices lost contact with each other, their context information was removed
from the local CiBs. As a result, when Device 1 joined into Device 0’s communication range
for the second and third time, more contextual packets had to be exchanged for their
registration in each other again.

Figure 7. Results of Scenario 2. Six SHARE packets were captured. The lease
timer was set to 1 second.

29
International Journal of Grid and Distributed Computing
Vol. 5, No. 4, December, 2012

When the lease timer was set to 2 seconds, the first leaving period (from 7.125 to 7.875
seconds in Figure 8) was ignored by Device 0 and Device 1. This is because during this
period, the devices only checked their lease timer once (at 7.5 seconds), so that when Device
1 rejoined at 7.875, both devices still had neighbors’ context information. Therefore, there are
no SHARE packets generated during the first and third joining periods.

Figure 8. Results of Scenario 2. Four SHARE packets were captured. The lease
timer was set to 2 seconds.

When the lease timer is set to 4 seconds, according to Table 5.9, the lease timer is longer
than both leaving periods, i.e. from 7.125 to 7.875 seconds and from 9.375 to 10.625 seconds.
Therefore, there is only one pair of SHARE packets being transmitted, which is when the two
devices communicate with each other at 6.5023 seconds.

Figure 9. Results of Scenario 2. Only two SHARE packets were captured. The
lease timer was set to 4 seconds.

According to the simulation results in the second scenario, the longer the lease timer, the
less contextual packets are generated. When the lease timer is set as 4 seconds, no rejoining
communication is needed since the context information of each device is kept until both
devices meet again.
Raising the lease timer in the CiComm protocol can reduce the number of redundant
contextual packets that result from frequent leaving and rejoining events. However, a trade-
off is required to set the lease timer since a large lease timer can also bring other issues. For
example, if the lease timer is extended, the updating of context information will be delayed,
which hence renders it out-of-date. In addition, it also causes more CiB resources to be
wasted through a reliance on out-of-date context. In the remaining simulations the lease timer
is set to be twice as long as the HB timer.

5. Conclusion
In this paper, a new approach to information management of mobile devices in a MANET
is discussed. The approach uses a context-aware framework to represent, store and exchange
relevant information of mobile devices, where dead loops and rejoining issues are effectively
eliminated. This is mainly achieved by setting up appropriate heart-beat timer and lease timer.
However, a trade-off is required to set the lease timer since a large lease timer can also bring
other issues. The update of context information can be delayed with the extension of the lease
timer, which hence renders it out-of-date.

30
 International Journal of Grid and Distributed Computing
Vol. 5, No. 4, December, 2012

Acknowledgements
This work was supported by A Project Funded by the Priority Academic Program
Development of Jiangsu Higher Education Institutions. It was also funded by The Project
Sponsored by the Scientific Research Foundation for the Returned Overseas Chinese Scholars,
Nanjing.

References
[1] A. Saeed and T. Waheed, “An Extensive Survey of Context-aware Middleware Architectures”, 2010 IEEE
International Conference on Electro/Information Technology (EIT), (2010), pp. 1-6.
[2] X. X. Diao, “Cooperative Inter-vehicle Communication Protocol Dedicated to Intelligent Transport Systems”,
in New Technologies, Mobile and Security, (2008), NTMS‟08.
[3] K. M. Hou, “LiveNode: LIMOS versatile embedded wireless sensor node”, Journal of Harbin Institute of
technology, vol. 32, (2007), pp. 139-144.
[4] R. Eigner and C. Mair, “Using Context Ontologies for Addressing and Routing in Mobile Ad Hoc Networks”,
in International Conference on Networking, (2009).
[5] K. Yang, “Context Aware Vertical Soft Handoff Algorithm For Heterogeneous Wireless Networks”, in
Vehicular Technology Conference, (2008), VTC 2008-Fall, IEEE 68th.
[6] S. Moon, “Visualization and Representation of Mobile Network Users, in Sensor, Mesh and Ad Hoc
Communications and Networks”, 5th Annual IEEE Communications Society Conference on SECON‟08,
(2008).
[7] A. Dey, “Understanding and Using Context”, Personal Ubiquitous Comput., vol. 5, Issue 1, (2001), pp. 4-7.
[8] M. Perttunen, J. Riekki and O. Lassila, “Context Representation and Reasoning in Pervasive Computing: a
Review”, International Journal of Multimedia and Ubiquitous Engineering, vol. 4, no. 4, (2009).
[9] M. Baldauf, S. Dustdar and F. Rosenberg, “A survey on context-aware systems”, International Journal of Ad
Hoc and Ubiquitous Computing, vol. 2, no. 4, (2007), pp. 263-277.
[10] M. D’Angelo, “A Layered Approach to Modelling and Design of Cross-Layer Protocols in Ad-Hoc Wireless
networks”, in Wireless Communications and Networking Conference, (2008), WCNC 2008, IEEE.
[11] A. A. Bakar, “Forming Trust in Mobile Ad Hoc Network”, in Communications and Mobile Computing,
(2009), WRI International Conference on CMC‟09.
[12] B. Garbinato, “Six-Shot Broadcast: a context-aware algorithm for efficient message diffusion in MANETs”,
Lecture Notes in Computer Science, (2008).
[13] S. Buchegger and J. -Y. Le Boudee, “Self-policing mobile ad hoc networks by reputation systems”,
Communications Magazine, IEEE, vol. 43, no. 7, (2005), pp. 101-107.
[14] M. Musolesi, “Adaptive routing for intermittently connected mobile ad hoc networks”, in World of Wireless
Mobile and Multimedia Networks, (2005), Sixth IEEE International Symposium on WoWMoM 2005.
[15] M. Musolesi and C. Mascolo, “Evaluating Context Information Predictability for Autonomic
Communication”, in World of Wireless, Mobile and Multimedia Networks, (2006), International Symposium
on WoWMoM 2006.
[16] R. Handorean, “Context aware session management for services in ad hoc networks” IEEE International
Conference on Services Computing, (2005).
[17] K. C. Lee and H. Cho, “Designing a Ubiquitous Decision Support Engine for Context Prediction: General
Bayesian Network Approach”, International Journal of u- and e- Service, Science and Technology, vol. 3, no.
3, (2010), SERSC.
[18] O. Riva, “Context-Aware Migratory Services in Ad Hoc Networks”, IEEE Transactions on Mobile
Computing, vol. 6, no. 12, (2007), pp. 1313-1328.
[19] J. Yamamoto, H. Nakagawa, K. Nakayama, Y. Tahara and A. Ohsuga, “A Context Sharing Message Broker
Architecture to Enhance Interoperability in Changeable Environments”, In: Mobile Ubiquitous Computing,
Systems, Services and Technologies, (2009), pp. 31-39.
[20] S. L. Kiani, M. Knappmeyery, N. Baker and B. Moltchanov, “A Federated Broker Architecture for Large
Scale Context Dissemination”, In: Computer and Information Technology (CIT), (2010), pp. 2964-2969.
[21] R. P. Pinto, “A Component Framework for Context-Awareness”, in Wireless Communications and Mobile
Computing Conference, (2008).
[22] D. Dudkowski, “Design and Implementation of a Reference Modelfor Context Management in Mobile Ad
Hoc Networks”, in Proceedings ofthe 22nd International Conference on Advanced Information
Networkingand Applications, (2008).
[23] Y. Chen, “LMSS A location-based multimedia service systemusing wireless networks”, in 8th International
Conference of IntelligentSystems Design and Application, (2008).

31
International Journal of Grid and Distributed Computing
Vol. 5, No. 4, December, 2012

[24] T. Ma, J. Huang and G. Wu, “An Architecture and Supporting Environment of Service-Oriented Computing
Based-on Context Awareness”, International Journal of Grid and Distributed Computing, vol. 2, no. 4, (2009).
[25] A. N. Mian, R. Baldoni and R. Beraldi, “A Survey of Service Discovery Protocols in Multihop Mobile Ad
Hoc Networks”, Pervasive Computing, IEEE, vol. 8, Issue 1, (2009), pp. 66-74.
[26] C. Cho and D. Lee, “Survey of Service Discovery Architectures for Mobile Ad Hoc Networks”, Term Paper,
Department of Computer and Information Science and Engineering (CICE), University of Florida, (2005).
[27] U. C. Kozat and L. Tassiulas, “Service Discovery in Mobile Ad Hoc Networks: an Overall Perspective on
Architectural Choices and Network Layer Support Issues”, Ad Hoc Networks, vol. 2, issue 1, (2004), pp. 23-
44.
[28] A. El Saoud, M. T. Kunz and S. Mahmoud, “SLPManet: Service Location Protocol for MANET”, in
Proceedings of the International Wireless Communications and Mobile Computing Conference IWCMC 2006,
(2006), pp. 701 -706.
[29] M. Abou, “BENCHManet: An Evaluation Framework for Service Discovery Protocols in MANET”, in
Sensor and Ad Hoc Communications and Networks, (2006), SECON 2006, pp. 860-865.
[30] Q. Liu, N. Linge and F. Gao, “A Context-aware Protocol for Location Discovery Exchange in a Ubiquitous
Environment”, International Journal of Future Generation Communication and Networking, vol. 7, no. 2,
(2012), pp. 177-182.

Authors

Qi Liu
BSc (2002), MSc (2006), PhD (2010) is an appointed Professor at the
Nanjing University of Information Science and Technology. His research
interests include context-awareness, data communication in MANET and
WSN, smart home and energy efficiency. He also devotes time to WSN
solutions on intelligent agriculture in protected field. He is a member of
IEEE and ACM.

Nigel Linge
BSC (1983), PhD (1987) is a Professor of Telecommunications at the
University of Salford. He is an electronic engineer by profession and has
research interests in network design, context aware information systems,
communication protocols, mobile systems and practical applications of
networking technology in areas such as energy and building monitoring.
He also devotes time to the preservation and dissemination of the history
of telecommunications and provides external advice to organization on
their heritage strategies. He is a Fellow of the IET and BCS.

32