Multiple time-scale energy scheduling with energy harvesting aided heterogeneous cloud radio access networks

In the next generation wireless communication networks, a large portion of operational expenditures (OPEX) has to be paid for energy consumption. Millions of base stations consume more energy in order to support increasing dramatically data traffic and diverse application services. Energy harvesting is becoming a promising technology to minimize the traditional power demand and thus reduce the OPEX. However, energy management and scheduling in conventional heterogeneous cellular network is greatly challenged. In this paper, we first introduce a hierarchical framework for the design and analysis of heterogeneous radio access network (H-C-RAN) with energy harvesting, which is composed of heterogeneous internet service provider (ISP) and remote radio heads (RRH) powered by energy harvesting facilities. The proposed hierarchical framework is feasible and tractable to utilize a market economical model which can work efficiently by utilizing energy sharing and energy trading among different ISPs. Moreover, we design user association scheme for D2D-assisted based on energy storage capacity and traffic load status of each small cell. Inspired by the fact that the energy harvesting rate, selection of user association and channel state vary in different time scale, we also propose a multiple time-scale energy scheduling strategy by modeling this issue using multitime scale Markov decision processes (MMDP). Simulation experiments indicate that the proposed strategies are able to significantly reduce the energy cost.