Adaptive congestion estimation and control in active queue management

Active Queue Management (AQM) policies provide an early indication of incipient congestion to the source. In this paper, we propose a new AQM policy that predicts the instantaneous queue length at next time instant using adaptive filtering technique. To use the algorithm in fast routers, we reduce its computational complexity using a direct method to compute the manipulating signals. We show that adaptive congestion estimation is able to control the oscillations in the instantaneous queue length. We compare the performance of our method with respect to other well-known AQM methods such as RED and PI which are also simulated by MATLAB. We also compare the computational complexity of these algorithms with each other.