Dynamic user association and energy control in cellular networks with renewable resources

In recent years, there has been a growing interest in equipping base-stations with renewable resources and batteries, with an aim to reduce the energy-related operational cost for cellular networks. To fully leverage the cost-saving potential of such base-stations, the design of joint resource-allocation and energy-control algorithms that take both the traffic-level dynamics and energy-level dynamics into account is called for. In this work, we propose a joint user-association and energy-control algorithm, where the induced operational cost can be made arbitrarily close to optimality at the expense of increased battery capacity in each base-station. The key to this algorithm is that it tries to match the traffic-profile with the energy-profile of the network, which leads to an efficient utilization of the renewable resources across the network. Simulation results show that a significant reduction in energy cost can be achieved by using our algorithm, even at a relatively small battery capacity.