Abstract: A dynamic energy management scheme of the energy harvesting base station powered by the smart grid with time-varying price is proposed. Specially, it was assumed that the base stations(BSs) were equipped with energy harvesting device, and the free energy collected by this device from renewable sources were stored in battery for BSs to use in future. Due to the randomness of energy harvest, smart grid was taken as a supplement of energy to ensure the stable operation of BSs. Based on queuing theory and Lyapunov optimization method, a dynamic energy management algorithm was proposed for BSs under two conditions of inelastic energy demand and elastic energy demand. Based on the proposed algorithm, BSs can dynamically choose to purchase energy from smart grid at low price and store it in battery for using at high price for reducing the energy cost of BSs. The proposed dynamic energy management algorithm has low complexity and does not need prior statistical information of energy collection, energy demand and time-varying price. The theoretical analysis showed that the proposed algorithm performed arbitrarily close to the optimal objective value, meanwhile, it ensured that the time delay did not exceed the tolerable time. Finally, simulation results showed the validity of the algorithm. And the effect of battery capacity on the performance of the algorithm was analyzed.

Key words: dynamic energy management, time-varying price, Lyapunov optimization, energy harvesting