Adaptive Step Size Control for Update of Activation Probability in Online Probabilistic BS Activation Control Method

We propose a new adaptive control method for the step size in updating the activation probability in an online probabilistic activation control method previously proposed by our group for pico-base stations (BSs) in heterogeneous networks where low transmission-power pico-BSs are overlaid onto a high transmission-power macro-BS. In the previous method, each pico-BS independently and iteratively updates the activation probability based on the time variation of the network-level system throughput, which is shared by BSs using the inter-BS information exchange, and the transition results of activated or deactivated states at each pico-BS. However, the previous work assumed a fixed step size in updating the activation probability. Therefore, there is room to improve the convergence property of the iterative algorithm. The proposed method adaptively changes the step size when updating the activation probability depending on the time variation in the system throughput levels. The proposed method brings about improvement in the convergence rate of the iterative algorithm and increases the system throughput after convergence. We show the effectiveness of the proposed method quantitatively by computer simulation.