Radio resource allocation in heterogeneous wireless networks: A spatial-temporal perspective

We study optimal radio resource allocation across multiple tiers of a heterogeneous wireless network in order to maximize the downlink sum throughput. Different from prior works, we consider both the randomness of base stations in space and dynamic user traffic session arrivals in time, accounting for both elastic and inelastic user traffic. A new stochastic analysis framework, which accommodates both spatial and temporal dimensions, is proposed to quantify the throughput objective. The derived throughput function is not in closed form and is non-concave in terms of the radio resource allocation factors to be optimized, hindering the search for an efficient optimization solution. Therefore, we further develop closed-form concave bounds that envelop the throughput function, to form convex approximations of the original optimization problem that can be solved efficiently. We characterize the performance gap when these bounds are used instead of the original objective. Both analytical bounding and simulation experiments demonstrate that the proposed solution is nearly optimal.