The power of online learning in stochastic network optimization

In this paper, we investigate the power of online learning in stochastic network optimization with unknown system statistics a priori. We are interested in understanding how information and learning can be efficiently incorporated into system control techniques, and what are the fundamental benefits of doing so. We propose two Online Learning-Aided Control techniques, OLAC and OLAC2, that explicitly utilize the past system information in current system control via a learning procedure called dual learning. We prove strong performance guarantees of the proposed algorithms: OLAC and OLAC2 achieve the near-optimal [O(ε), O([log(1/ε)]²)] utility-delay tradeoff and OLAC2 possesses an O(ε^-2/3) convergence time. Simulation results also confirm the superior performance of the proposed algorithms in practice. To the best of our knowledge, OLAC and OLAC2 are the first algorithms that simultaneously possess explicit near-optimal delay guarantee and sub-linear convergence time, and our attempt is the first to explicitly incorporate online learning into stochastic network optimization and to demonstrate its power in both theory and practice.