New scheduling policies for multiclass queueing networks: applications to peer-to-peer systems

We propose a new parametric class of scheduling policies for closed multiclass queuing networks that steer instantaneous throughputs towards a fixed target. These policies measure distance from the target using a weighted norm. We establish that with proper target selection the corresponding policy is efficient, that is, attains bottleneck throughput in the infinite population limit. The proposed policies are amenable to distributed implementation using local state information. We exploit our earlier work in [D. Bertsimas et al., 1994] to select appropriate parameter values and outline how optimal parameter values can be computed. We report numerical results indicating that we obtain near-optimal policies. Finally, we consider an application of our work to Internet file-sharing peer-to-peer systems and show that our policies can lead to significant performance improvements.