Project 3-4: Robust Wireless Communications in Complex Environments

Principal Investigators: Richard Dean, Gregory Wilkins, and Yacob Astatke (Morgan State University)

A warfighter’s ability to function in a complex computational battlefield environment is complicated by the mix of tactical and strategic assets that are constantly changing as he moves through the environment. Much is known today about standalone wireless mobile ad hoc networks (MANET) and much is known about fixed network resources. Little is known, however, about how these resources can be dynamically mixed to support the mission.

AHPCRC researchers are focusing on the development of a network strategy that mixes fixed and mobile networks to the overall performance advantage of both. This work draws on multistage clustering techniques that have the potential for near optimum performance with an inherently efficient computational scheme.

The Mixed Network solution developed at Morgan State University addresses the performance of fixed telemetry ground networks with associated mobile nodes (test articles) within the range of the ground station and disconnected mobile nodes that are over the horizon and operate as a MANET.  A multistage clustering scheme can organize all the nodes into clusters that are in either a ground-based cluster or in one of  several ad hoc clusters. Such a scheme can jointly enhance the performance of all the nodes using a distance measure that includes location, Quality of Service performance, and interference management of these nodes in static and dynamic environments.

Research will focus on extending the current Mixed Network strategy to a larger framework and exploring the computation and performance features. Such a strategy would move from a single fixed network with multiple ad hoc networks to a multiplicity of fixed networks with a multiplicity of associated ad hoc networks. Nodes could gravitate toward competing fixed network elements on the basis of service and performance. "Market behavior" captured in distance measures would adapt network traffic to support best-value services.