CONGESTION CONTROL OF DATA NETWORKS USING ROBUST CONTROLLER

In this article a new method for queue management has been proposed using robust control. In computer networks and routers different reasons including: speed variation in output, delay and losses causing congestion. One of the queue management methods to solve congestion problems is active queue management controlling queue size through changing probability of missing data and getting feedback from queue size. In this method, a new state variable, which is the error between queue reference size as well as real queue size, has been augmented to state space model. The robust controller has been designed for the new model using H∞ control. Therefore, the close loop system is stable and robust in facing with variations of the parameters and conditions such as output speed, input speed, delay and other disturbances. In addition to robustness, the queue size follows the reference queue size in the presence of the disturbances thank to the new state variable.

Keywords:

Congestion Control , Active queue management (AQM) , Robust Control , H∞ control