Recognizing that Internet congestion control is a complex nonlinear system, we propose here to use an intelligent controller to improve its stability and performance. In particular, we propose here a new, powerful, easy-to-configure and robust active queue management (AQM) scheme called adaptive neuron AQM (AN-AQM). We present extensive simulation results for AN-AQM, over a wide range of network conditions and scenarios, that demonstrate its attributes. We demonstrate its robustness in various realistic environments involving bursty HTTP connections and non-responsive UDP connections. Comparison with other AQM schemes has demonstrated the superiority of AN-AQM over well-known AQM schemes in achieving faster convergence to queue length target, and smaller queue length jitter. © IFIP International Federation for Information Processing 2007.
CITATION STYLE
Sun, J., & Zukerman, M. (2007). An adaptive neuron AQM for a stable internet. In Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics) (Vol. 4479 LNCS, pp. 844–854). Springer Verlag. https://doi.org/10.1007/978-3-540-72606-7_72
Mendeley helps you to discover research relevant for your work.