Implementation issues of a multi-class flow control algorithm

Various implementations that could support the optimal flow control algorithms using the Pareto optimal criterion are explored. The performance of the Pareto algorithms for flow control is analyzed. Simulation results suggest that the convergence is better when the observation windows are maintained at the node level and are packet count based. The window size has to be sufficiently long. The relaxation techniques are useful with the right choice of relaxation factors. The greedy algorithms which use estimation techniques to achieve complete decentralizability are found to be very appealing.<
>