AOPC: An Adaptive Optimized Proportional Controller for AQM

Jianxin Wang,Liang Rong
DOI: https://doi.org/10.1109/ICPP.2006.25
2006-01-01
Abstract:In this paper, we propose a novel AQM algorithm based on the optimized second-order system model for the first time. The algorithm is called Adaptive Optimized Proportional Controller (AOPC). The system design of AOPC is based on classical TCP/AQM interconnection system model, which has a strong relation to the network load level. Through introducing network load estimator in the TCP/AQM interconnection system model, AOPC is capable to detach from the number of TCP sessions N and insensitive to varying network conditions. Meanwhile, AOPC surmounts the parameter configuration difficulties experienced by many AQM algorithms. The parameter tuning rule is compliance with the optimized second-order system model which results fast convergence rate. The performances of AOPC are compared to REM, PI, PID, PIP, and LRED using NS2 simulations. From the experiments and analysis, we conclude that AOPC outperforms other AQM algorithms that it obtains stable queue evolution, achieves the fastest convergence rate to equilibrium point in time-varying network conditions than other algorithms, and fulfills perfect tradeoff between link utilization and queueing delay.
What problem does this paper attempt to address?