Congestion Control in Mobile Networks

Current research is focused on adaptive resource management protocols to offer quality of service guarantees within mobile networks. The usual approach to protocol evaluation is statistical analysis of simulation results. This is insufficient for understanding complex behaviors of distributed, adaptive algorithms. We propose adaptive admission control algorithms to manage mobile network congestion caused by failures, and data visualization for algorithm analysis. Examples illustrate the dramatic impact of even simple visualizations on enabling the protocol designer to clearly understand real-time algorithm behavior under highly variable conditions.