ADAPTATION OF THE APPLICATION OF SMITH'S PREDICTOR FOR CONTROLLING THE VOLUME OF BUFFER MEMORY IN INFORMATION NETWORKS WITH RANDOM DELAYS

Abstract

This article examines the issue of congestion control in a general network with packet switching, using classical control theory. Since data propagation delays play a key role in high-speed communication networks, the Smith predictor was chosen to develop a simple congestion control law that is effective in smoothing load shedding and reducing throughput delay.  The behavior of the control system depends on the rate of change of the input data flow. The available "BEST EFFORT" bandwidth is modeled as an unknown and limited disturbance input because it is difficult to measure.