- Biomedical Optics & Medical Imaging
- Defense & Security
- Electronic Imaging & Signal Processing
- Illumination & Displays
- Lasers & Sources
- Micro/Nano Lithography
- Optical Design & Engineering
- Optoelectronics & Communications
- Remote Sensing
- Sensing & Measurement
- Solar & Alternative Energy
- Sign up for Newsroom E-Alerts
- Information for:
Optoelectronics & Communications
A few powerful nodes enhance mobile network connectivity
Strategically positioned nodes with extra transmission power can maintain system-wide connectivity in ad hoc networks.
6 September 2006, SPIE Newsroom. DOI: 10.1117/2.1200608.0360
A mobile ad hoc network (MANET) is a temporary collection of wireless nodes communicating without the benefit of infrastructure. A connection between two nodes may involve several others in what is known as multi-hop routing. MANET can be used in situations in which, for example, soldiers relay and share information on a battlefield or relief workers coordinate efforts during an emergency.
A basic requirement of a MANET is that it must maintain higher network connectivity, loss of which could entail severe consequences. Therefore, a principal task is to set up a route that connects source and destination nodes. Due to the dynamic shifts in network topology, multi-hop routing is neither stable nor pre-established.
Several different MANET routing protocols have been proposed in recent years,1–4 but research has shown that they are incapable of maintaining high connectivity in the face of diminished node density.5,6 Our simulation7 confirms this. Moreover, the protocols may require bidirectional links such that if node A can receive data from node B, then B can also receive data from A. However, this is not always the case. There might be numerous unidirectional links where node transmission power is affected by interference, building blockages, etc.
Our approach to enhance MANET network connectivity is to employ a few nodes with larger transmission power as strategic or backbone nodes (BN). These help to establish connections between remote nodes that are otherwise unable to connect. In Figure 1, BN1 and BN2 are the strategic nodes with the capacity for mesh connecting with each other. Nodes in the upper area (A, B, and G) are sufficiently far from nodes in the lower area (C, D, E, and F) that no node in the lower area could directly reach any node in the upper area, and vice versa. With the assistance of BN1 and BN2 however, such connections can be established.
Figure 1. In this model of backbone-node (BN)-supported wireless ad hoc networks, nodes A, B, and G connect to C, D, E, and F via BN1 and BN2.
Routes are set up as follows: mobile nodes near each other can have direct connections (e.g. C → D), and mobile nodes near BNs but without direct connections can be connected via BNs (e.g. E → BN2 → C → D). Mobile nodes between remote areas can also be connected via BNs (e.g. A → B → BN1 → BN2 → C → D). Thus, whenever a remote mobile node has a request to connect to another node, the BNs function both as normal mobile nodes and also forward connection requests to the desired remote destination node. This implies that the networks can establish more and better routes, and are able to deliver more packets without extra routing overhead.
The effectiveness of this approach in enhancing network connectivity can be proved by simulation in different network configurations, that is, with different node densities, velocities, and transmission power. We used the Java-based scalable wireless ad hoc network simulator (SWANS) to simulate a MANET with IEEE 802.11b medium access control protocol and transmission of UDP traffic. The BNs possess all the properties of normal mobile nodes, but they possess larger transmission power and remain stationary during the simulation. The ad hoc on-demand vector routing (AODV) protocol is used. Two criteria are used to assess network connectivity: the packet delivery ratio (the total number of data packets received over the total number of packets sent by all nodes) and the average path length (the average number of mobile hops a packet takes to reach its destination).
Results show that our approach performs better than the normal MANET in all simulated network scenarios. Performance is significantly improved (see Figure 2) when network node density is lower. In the case of 50 nodes, our approach can deliver more than 89% of packets, an increase of 28% compared with the normal MANET. This is because, without the assistance of BNs, many nodes are out of the reach of each other when the node density is lower. In comparison with the normal network, the average packet length is shorter, with our network consistently establishing routes about one hop less than the normal network in all scenarios.
Figure 2. In comparing packet delivery ratios as a function of node density when nodes are stationary, the network with BNs provides better performance, especially in lower node density situations.
High connectivity, extremely important for MANET to be feasible in critical situations, can be greatly enhanced by selecting a few nodes to serve as backbone nodes. This approach is relatively easy to deploy inasmuch as output power of many wireless devices can be custom-configured. It can also be used to improve user or node connectivity in other networks, such as sensor networks and mobile communication systems.
Xiyu Shi, Ahmet Kondoz
Centre for Communication Systems Research, University of Surrey
Dr X. Shi is a research fellow at the Centre of Communication Systems Research in the University of Surrey, UK. His research interests include mobile network routing, QoS, network security, and distributed virtual reality.
Ahmet Kondoz is professor and deputy director of the Centre of Communication Systems Research in the University of Surrey, UK. His current research interests are low bit rate speech, image and video coding error-resilient video transmission, mobile multimedia communications, robust wireless ATM, real-time terminal design and implementation for mobile communications. He is the author or co-author of more than 130 publications.
Department of Information Systems, University of Buckingham
Christopher Adams is professor at the Department of Information Systems in the University of Buckingham, UK. He holds a join appointment with the Rutherford Appleton Laboratory, UK. His research interests are the performance of operating systems, scalability of high performance network and network components, and real-time video services in network environments.