Load Based Dynamic Small Cell On/Off Strategy In Ultra-Dense Networks

Abstract

With the deployment of small cells (SCs) becoming denser, the power consumption will increase sharply. How to minimize the power consumption of the network attracts widespread concern. Most researches adopt heuristic algorithms to switch the small cells, but few consider that the neighbor cells' load will affect the switching process which may make the algorithm terminate prematurely. Moreover, when a cell's state changes, there is a switching energy cost incurred by changing users attachment. It is a significant amount of power consumption and cannot be ignored. However, it's hard to optimize the switching cost in the switching process. Therefore, we minimize the total power consumption by two steps. Firstly, aiming to decrease the switching cost, we propose a centralized user association strategy to make the users served by the cell having lower probability of being closed. Secondly, in order to further reduce energy consumption and overcome the weakness of existing heuristic algorithm, we propose a neighbor cells' load based switching algorithm to make the switching off order more reasonable. Simulation results demonstrate that our association strategy reduces switching cost significantly and our cell switching algorithm will save more power than existing heuristic algorithms. The total energy can be saved up to 52%.