The Study Of Energy Efficiency Optimization Algorithm Based On Coordinated Multi-point Transmission

With the progress of communication technology and the growing tend of energysources shortage, optimization of energy utilization and the sustainable development havebecome one of the hot topics in academia. Base station costs much energy consumption inwireless cellular communication networks, which causes huge waste in the case of thenumber of users reduces to little. It goes against the energy optimization and theimplementation of environmental protection. Therefore, it is necessary to reduce theenergy consumption of base stations in order to optimize the energy efficiency of wirelesscellular network.In allusion to the energy efficiency optimization, base station sleeping strategy isproposed, in which the base station should be turned off when the traffic is low and thesurrounding base stations cooperate to serve the users. It combines the coordinatedmulti-point communication and base station sleeping strategy in order to implement anoptimal CoMP grouping to sever the sleeping cell. Concrete research content is as follows:Firstly, set up a19-cell model of communication system, implement the base stationsleeping strategy in the center cell, and set the first layer of cellular cells as a CoMP setwhile the second layer as a noise source, and apply large-scale fading model andsmall-scale fading model transmission model.Secondly, choose the different optimization algorithm according to the number ofusers in the base station sleeping strategy. The threshold method is proposed to optimizethe energy efficiency and the downlink performance for each single user when the numberof them is low in the sleeping cell. When it comes to more users，Q-learning algorithm ofthe reinforcement learning is used to find an optimal CoMP grouping for the wholesleeping cell. With the downlink performance acted as the enhanced feedback, the optimalstrategy enhances the cell energy efficiency as well as insures the communication qualityfor each user. And take one step Q-learning algorithm and multi-step Q-learning algorithmof efficiency optimization for stationary and mobile users, respectively.Finally, simulate the implemention of the base station sleeping strategy and switchbetween optimization algorithms according to the number of users changing within a day. Simulations verified the optimization algorithm enhances both the energy efficiency andthe downlink performance, which makes the cellular network become more energy-saving,efficient and environmental.