Performance Modeling of Mobile Ad Hoc Networks

Ognjen Zivojnovic

Wireless ad hoc networks of mobile devices have the potential to provide full connectivity with minimal infrastructure and deployment time required. Furthermore, these networks would theoretically be self-adapting to geographic bandwidth requirement distributions, eliminating the need for careful planning of base station locations. Due to their flexibility, self-sufficiency and low installation overhead, ad hoc networks are a promising candidate for the future of wireless network systems.

There are still very large technical challenges to be solved before this type of network can truly become practical. Amongst other issues, two major challenges present are the power requirements placed on individual cell phones or other mobile wireless devices by such a network, as these now need to not only communicate their own data, but also act as a router for other, nearby, devices. The routing in on itself presents further challenges, as data needs to be routed through a set of nodes (other mobile devices on the network) between the destination device and a source transmitter. This is challenging as one has to optimize both the overall bandwidth utilization amongst all devices present, as well as the delay due to the routing of data designated for a specific destination. The routing has to be performed without global knowledge of the network topography, with each router instead depending only on information passed on to it from individual, adjacent, nodes.

In my project I will attempt to model these challenges. I plan to execute the project in two phases. First, I intend to create a Matlab framework, that, given a device density, a path loss and fading model, and the power transmitted by each device, will calculate the power received by every other device in a random instance of the scenario, which each device having a randomized location and velocity. I will also enable this model environment to develop over time, enabling the observation of the changing signal levels between nodes as the individual nodes continuously change positions due to their randomized velocities (with the velocities being modeled as Markov chains).

Following the completion of this framework, I plan to research ad hoc routing algorithms, and implement them into my model. Ideally the final product would enable a user to select a channel model and a routing algorithm, and then observe a time varying map of the interconnectivity between devices to gain a general idea of how effectively the network is connected. The program would also keep track of what percentage of devices would be experiencing outage and the average transmission delay (assumed to be proportional to the number of edges that need to be transversed between the source and destination node), in order to enable judging of the performance of a routing algorithm given a channel, device density and power limitations.

Informal list of sources

our textbook

http://www.open-mesh.net/

Ozan, K. Tonguz, Gianluigi Ferrari (May 2006). John Wiley & Sons.. ed. Ad Hoc Wireless Networks: A Communication-Theoretic Perspective.

C K Toh (January 2002). Prentice Hall Publishers. ed. Ad Hoc Mobile Wireless Networks: Protocols & Systems.

Royer, E., Toh, C. (April 1999). "A Review of Current Routing Protocols for Ad Hoc Mobile Wireless

Networks". IEEE Personal Communications

Mauve, M., Widmer, J., Hartenstein, H. (December 2001). "A Survey on Position-Based Routing in

Mobile Ad Hoc Networks". IEEE Network