Connectivity management in mobile ad hoc networks using particle swarm optimization
Introduction
Mobile ad hoc networks (MANET) are formed without a central administration so that nodes transmit packets on behalf of other nodes. The most commonly cited application of MANET is military communications including combat, emergency response, and search/rescue maneuvers [1], [2], [3], [4], [5].
There are many factors that affect the performance and reliability of a MANET. Links between the mobile devices sometimes exist, and sometimes do not, depending on the devices’ locations relative to each other, their transmission power and the surrounding environment. New mobile devices can enter the system or existing devices can leave for various reasons including loss of battery power or loss of signal strength (due to distance or environmental causes). Assuming random user behavior, it is very likely that one or more users will lose connectivity with the network or with the parts of the network due to their positions relative to other users. If a user is outside the range of its nearest neighbor in terms of signal strength, then its access to the rest of the network is unavailable.
This paper proposes a dynamic MANET management system to maintain the connectivity of a MANET by using controlled ad hoc network nodes (called agents), where the global state of the network can be tracked using a Global Positioning System (GPS). In a GPS, each node is equipped with a receiver and the location data is periodically transmitted to a central location using low frequency radio or a satellite modem embedded in the node [6]. In this paper, user nodes are allowed to move freely and their current and past location data are available by the use of the GPS and a kinematic-based prediction model used to estimate their future locations. The purpose of agents is to augment network connectivity. Agents are moved each time step during the operation of the network to optimize network connectivity among user nodes. (Note that networks operate in continuous time, however to make the optimization tractable the continuous time operation is sliced into discrete time steps.) The predicted locations of users are used in a particle swarm optimization (PSO) algorithm to dynamically direct the motion of the mobile agents. Among application areas of the proposed approach are network-centric warfare military, fleet tracking, and search and rescue operations, where network nodes have power sources to support a GPS.
There has been very limited work in the literature to improve network connectivity through mobile agents. In this paper, multiple mobile agents are considered and no restrictions are imposed on the movement of user nodes. The proposed approach is applicable to all user movement models and scenarios. The objective function is a network connectivity measure using a maximum flow formulation that aims to increase overall network performance, independent from the routing protocol used. Therefore, the problem is studied for a much more general case than previously published work.
Section snippets
Background
In the literature, several approaches have been proposed to address the challenges in MANETs due to unpredictable user node movements. One of the major problems is the accessibility of the centralized network services used by all network nodes when the network is disconnected. A solution approach to this problem involves replicating network services [7], [8] and critical data [9] at multiple nodes and dynamically deploying these nodes to disconnected partitions of the network. Another problem
Assumptions and problem inputs
There are two main types of MANET nodes; user nodes and agent nodes. User nodes are the nodes that demand network service. Mobile agents are responsible for helping the user nodes experience the best network service possible. The user nodes in the MANET move at their own will and it is assumed that their future positions are unknown. However, the location data of users and agents are assumed to be available to the agent control system at all times. This is technically possible by using a GPS [6]
The particle swarm optimization
Particle swarm optimization (PSO), developed by Eberhart and Kennedy ([37]), is a population based optimization tool which emulates the social behavior of species that live in the form of swarms in nature. These swarms are capable of exchanging valuable information such as food locations in the habitat. Like a genetic algorithm, PSO has a population of randomly initialized candidate solutions. Different from evolutionary algorithms, swarm particles do not mate nor mutate to create offspring.
Computational experiments
Both dynamic and static problems were tested. The former covers many time steps of network operation and would be of the type used in practice. The latter was used to gauge comparison with a MIP approach and to evaluate scalability to larger sized problems. For static problems, the PSO algorithm stops and returns a solution after 1000 iterations and for dynamic problems the requirement for each time step is set equal to 100.
Conclusions
In this paper, a new model is proposed to conceptualize an autonomous topology optimization for mobile ad hoc networks using multiple mobile agents. The representation of wireless ad hoc network communications as network flows and optimization using a maximum flow model is a novel and advantageous approach. This representation is very responsive to small changes in topology when evaluating network connectivity and performance. Also, it can be used with any signal attenuation model when
Orhan Dengiz received his bachelor’s degree from Middle East Technical University (METU) in civil engineeinring in 2000. In August 2000, he joined Auburn University’s Department of Industrial and Systems Engineering as a graduate student. He received a master’s degree in 2002, and Ph.D. degree in 2007. Dengiz has worked on inter-disciplinary projects that involved materials science, computer and machine vision, and industrial engineering. In 2004, he spent one month at the Machine Vision
References (52)
Mobile ad hoc networking: imperatives and challenges
Ad Hoc Networks
(2003)- et al.
Data communications and topology control in wireless ad hoc networks
Ad Hoc Networks
(2005) - et al.
Communication in disconnected ad hoc networks using message relay
Journal of Parallel and Distributed Computing
(2003) Motion planning for cooperative unicycle-type mobile robots with limited sensing ranges: a distributed receding horizon approach
Robotics and Autonomous Systems
(2009)Collective robotic search using hybrid techniques: fuzzy logic and swarm intelligence inspired by nature
Engineering Applications of Artificial Intelligence
(2009)- et al.
A reliable on-demand routing protocol for mobile ad hoc networks with mobility prediction
Computer Communications
(2005) - et al.
Location tracking in mobile ad hoc networks using particle filters
Journal of Discrete Algorithms
(2007) - et al.
Multi-strategy ensemble particle swarm optimization for dynamic optimization
Information Sciences
(2008) A modified particle swarm optimization-based dynamic recurrent neural network for identifying and controlling nonlinear systems
Computers and Structures
(2007)Wireless ad hoc networking-The art of networking without a network
Ericsson Review
(2000)
Understanding GPS: Principles and Applications
Ant-based swarming with positionless micro air vehicles for communication relay
Swarm Intelligence
Cited by (59)
Testing AGV mobility control method for MANET coverage optimization using procedural simulation
2022, Computer CommunicationsCitation Excerpt :The communication is clustered per cooperatively working AGVs and orchestrated by SDN. For ad-hoc networks, there have also been works, that utilize the mobility of nodes in order to optimize the network topology [23,24]. It is quite common to plan the movement of nodes in Mobile Sensor Networks (MSNs) with the goal to establish communication [25,26].
A Dynamical Relay node placement solution for MANETs
2017, Computer CommunicationsCitation Excerpt :On the other hand, in dynamically changing scenarios, static solutions cannot be successfully applied because the location optimality at a certain time does not hold during network operation. Therefore, proposals addressing RN dynamic re-positioning according to the changes in the network topology are necessary [7–10]. Although dynamic RN re-location solutions provide more adaptable systems against different circumstances, they involve additional challenges to be addressed in comparison with their static counterparts.
Traffic and mobility aware resource prediction using cognitive agent in mobile ad hoc networks
2016, Journal of Network and Computer ApplicationsEvolutionary Algorithms for Enhancing Mobile Ad Hoc Network Security
2022, Internet of Things
Orhan Dengiz received his bachelor’s degree from Middle East Technical University (METU) in civil engineeinring in 2000. In August 2000, he joined Auburn University’s Department of Industrial and Systems Engineering as a graduate student. He received a master’s degree in 2002, and Ph.D. degree in 2007. Dengiz has worked on inter-disciplinary projects that involved materials science, computer and machine vision, and industrial engineering. In 2004, he spent one month at the Machine Vision Laboratory, University of the West of England, Bristol, UK.
Dengiz’s research involves application of computational techniques such as heuristic optimization algorithms and artificial intelligence tools in areas of manufacturing, system reliability design, automated vision systems, wireless telecommunication network design and automation problems. He published many articals in Materials Science and Engineering, International Journal of Production Research, Manufacturing, and Logistics Engineering, Computers in Industry, Journal of the European Ceramic Society and Journal of Operational Research Society besides many international Conference Proceedings. Publications.
After completing his doctoral studies, Dengiz returned to Ankara-TURKEY, and he has been working as a manager in a private group of companies operating in construction, energy and recycled synthetic textile production areas.
Abdullah Konak is an Associate Professor of Information Sciences and Technology at the Pennsylvania State University Berks. He received his degrees in Industrial Engineering, B.S. from Yildiz Technical University, Turkey, M.S. from Bradley University, and Ph.D. from the University of Pittsburgh. His current research interest is in the application of Operations Research techniques to complex problems, including such topics as telecommunication network design, network reliability analysis/optimization, facilities design, and data mining. He has published papers in IIE Transactions, Operations Research Letters, INFORMS Journal on Computing, OMEGA-The International Journal of Management Science, IEEE Transactions on Reliability, International Journal of Modeling and Simulation, International Journal of Production Research, Engineering Optimization, and Journal of Intelligent Manufacturing. He is a member of IIE and INFORMS.
Alice E. Smith is Professor and Chair of the Industrial and Systems Engineering Department at Auburn University. Previous to this position, she was on the faculty of the Department of Industrial Engineering at the University of Pittsburgh, which she joined in 1991 after industrial experience with Southwestern Bell Corporation. She has degrees in engineering and business from Rice University, Saint Louis University and Missouri University of Science and Technology.
She holds one US patent and several international patents and has authored more than 200 publications which have garnered over 1000 citations (ISI Web of Science). She won the E.L. Grant Best Paper Awards in 1999 and in 2006, and the William A.J. Golomski Best Paper Award in 2002. Several of her papers are among the most highly cited in their respective journals including the second most cited paper of IEEE Transactions on Reliability. She currently holds editorial positions on INFORMS Journal on Computing, Computers & Operations Research, International Journal of General Systems and IEEE Transactions on Evolutionary Computation.
She has served as a principal investigator on over $4 million of sponsored research. Her research in analysis, modeling and optimization of complex systems has been funded by NASA, NIST, DOT/FHWA, Lockheed Martin, Adtranz (now Bombardier Transportation), the Ben Franklin Technology Center of Western Pennsylvania and NSF, from which she was awarded a CAREER grant in 1995 and an ADVANCE Leadership grant in 2001. Her industrial partners on sponsored research projects have included DaimlerChrysler Electronics, Eljer Plumbingware, Extrude Hone, Ford Motor, PPG Industries and Crucible Compaction Metals. International research collaborations have been sponsored by the federal governments of Japan, Turkey, United Kingdom, the Netherlands, Egypt, South Korea, Iraq and the US and by the Institute of International Education.
She was awarded the INFORMS WORMS Award for the Advancement of Women in OR/MS in 2009. She was named an Auburn University Philpott-WestPoint Stevens Distinguished Professor in 2001, received the Senior Research Award of the College of Engineering at Auburn University in 2001 and the University of Pittsburgh School of Engineering Board of Visitors Faculty Award in 1996.
Five of her doctoral students are in tenured positions at US universities and two of these are NSF CAREER awardees. A further three doctoral students are tenured or tenure track faculty at foreign institutions. She is a fellow of the Institute of Industrial Engineers, a senior member of the Institute of Electrical and Electronics Engineers and of the Society of Women Engineers, a member of Tau Beta Pi, the Institute for Operations Research and Management Science and the American Society for Engineering Education, and a Registered Professional Engineer in Industrial Engineering in Alabama and Pennsylvania.
- 1
Tel.: +1 334 844 1400; fax: +1 334 844 1381.