Congestion Control in Multi-Source Communication Networks - a Time-Varying Sampling Period System Case Study

In this paper the problem of flow control in a multi-source connection-oriented network is addressed. In the considered network the feedback mechanism is provided by control units generated by each source once every M data packets. As the rate may be modified only, when the management unit arrives back at the source with the feedback information incorporated, the transfer speed is changed at irregularly spaced time instants. A new nonlinear strategy effectively combining the Smith principle with the proportional controller with saturation is proposed. Conditions for data loss elimination and full bottleneck link bandwidth utilization are presented and strictly proved with explicit consideration of the aperiodic feedback information availability. Since the controller saturation limits are fully adjustable, the algorithm performance may be easily tuned according to the existing system resources.