Asymptotically optimal priority policies for indexable and nonindexable restless bandits

I. M. Verloop

We study the asymptotic optimal control of multi-class restless bandits. A restless bandit is a controllable stochastic process whose state evolution depends on whether or not the…