A novel energy aware dynamic on-off control of base stations in wireless networks

To reduce network energy consumption without losing system performance is a key technique in 5th generation wireless communication systems. The on-off control in multi base stations scenario is proved a mixed dynamic integer programming problem and is hard to be solved. In this work, we provide a novel energy aware dynamic on-off control method in multi-base station scenario to tackle this problem. First of all, we group different base stations in a given area by utilizing the joint correlation of distance and traffic load under the consideration of ideal cooperation; then we perform on-off operations in every group by predicting the upcoming traffic load. Our novel method could reduce the degree of the optimization problem and be processed online with lower complexity. Simulation result indicates that, by using our novel method, we could reduce network energy consumption up to 23% compared to the reference method based on statistical on-off control.