An adaptive AQM algorithm based on neuron reinforcement learning
Chuan Zhou, Dongjie Di, Qingwei Chen, Jian Chun Guo · 2009
In recent years, it has become an active research direction to develop adaptive and robust active queue management (AQM) scheme for congestion control of complex time-varying network. A novel adaptive AQM scheme based on neuron reinforcement learning (NRL) is presented in this paper. This scheme uses queue length and link rate as congestion notification to determine an appropriate drop/mark probability, and the parameters of neuron can be adjusted online according to the time-varying network environment so that the stability of queue dynamics and robustness for fluctuation of TCP loads are guaranteed. This scheme is easy to implement with simple structure, and it is independent of the model of plant to be controlled. Simulation results show that this proposed algorithm is especially suitable for solving the complex network congestion control problem, and also has better stability and robustness.