Routing arrivals to M parallel queues

The authors have previously (1994) addressed the problem of finding the optimal sequence to route N calls to M queues in parallel, given the prior distribution on the number in system, service time, and interarrival time distributions. We present here several extensions to the above models and results: (a) we consider state dependent service rates; and (b) routing of multiclass arrivals. Furthermore, we discuss the extension of our modeling paradigm to an infinite horizon framework, both for the open-loop and the periodic updates scenarios