Operation Cost Minimization for Base Stations With Heterogenous Energy Supplies and Sleep-Awake Mode: A Two-Timescale Approach

In this paper, we study an energy cost minimization problem in cognitive mobile wireless networks, where base stations (BSs) are powered by hybrid energy sources including external power grids (PG), harvested renewable energy (RE), distributed local generators and energy storages. Based on sleep mode techniques, idle BSs can be switched to an energy-saving mode to cut down unnecessary expenditure. We formulate the problem into an optimization programming to achieve optimal decisions for energy scheduling and sleep control. By adopting a two-timescale approach, we implement the power scheduling and data transmission in time slots and the BSs sleep-awake mode in time frames (each time frame consists of multiple time slots) to avoid the overheads for frequent sleep/awake mode switches. Based on the Lyapunov technique, we further propose a close-to-optimal algorithm which only requires mean energy price of PG in each time frame instead of the exact future information about stochastic inputs (e.g., the amount of RE harvesting and user demand for data traffic). The proposed algorithm can achieve approximately minimal energy cost and ensure the stability of workload and battery queues. We present theoretical analysis as well as numerical simulations to demonstrate the performance of the proposed algorithm.