Periodic-Collaboration-Based Energy-Efficient Cell Dormancy in Heterogeneous Dense Networks

Abstract

In order to adapt to the change of mobile service needs, a base station sleep method combining long cycle and short cycle style is proposed to realize energy saving in heterogeneous dense networks. In the long-cycle mode, the neighboring base stations cooperate to optimize the activation and deactivation strategy. Meanwhile, a constrained graph game is formulated where each base station is abstracted as a game player with the constraint of traffic load. Then the base station states are solved distributively to reduce signaling overhead and base station power consumption. In the short-cycle mode, the transmission mode and non-transmission mode are introduced to enhance the adaptability to fast-changing services. The base stations are clustered and the mode matrix of the base station cluster is set to form the potential cooperation among base stations, thus energy saving is achieved by reducing inter-base station interference. Then, a distributed iterative algorithm for solving the generalized Nash equilibrium and the selection process of the mode matrix of the base station cluster is given. Simulation results show that the proposed method can respectively save energy by more than 12% and 15% during 24 hours, compared with the reference log-cycle and short-cycle modes.