Research On Intelligent Power Consumption Method Of Communication System In Renewable Energy Microgrid

Green communication is a hot topic in current wireless communication.Through the research of green communication,it can effectively reduce the energy consumption of power grid and improve the utilization rate of resources,which is of great significance to the construction of wireless communication at present and in the future.The research scope of green communication is very extensive.This paper focuses on the wireless communication system based on energy harvesting.Through effective power allocation and energy cooperation,the base station can make efficient use of energy.This paper firstly investigates the development of green communication and renewable energy microgrid.Secondly,the state of art of power allocation and energy sharing in communication systems is investigated,and the existing power allocation methods of communication systems are analyzed.In this paper,we aim to maximize the average throughput of the system and study the power allocation and energy sharing strategies in the scenario of two cellular base stations,multiple microcell base stations and large-scale cellular base stations.The specific research content is:(1)The power consumption methods of two energy harvesting cellular base stations are studied.We propose a power allocation and energy cooperation strategy based on Nash Q learning.The proposed algorithm does not need perfect information of the environment.(2)The user’s multi-association mode is considered,and the association problem of users,base stations and channels is modeled as a matching process.Regarding the base station and its channels as one matching object,and the matching problem of the three sets are simplified to two sets,and the multi-association model of the user is established.In the power allocation phase,the power allocation problem of multiple APs is modeled as a dynamic game,and the exploration phase is introduced.The algorithm learns a stable power allocation and energy sharing strategy in the exploration phase.(3)The online power allocation method in the large-scale cellular base station scenario is studied.When the number of microcell base stations tends to be infinite in an ultra-dense network,the traditional analysis methods become infeasible due to the existence of a dimensional disaster.In this paper,the mean field theory is used to model the interference problem in communication system,and the mean field approximation method is used to decouple the interference.Combined with the reinforcement learning algorithm,a data-driven MFG equation solving method is proposed.Finally,we give the simulation results of the above three research contents,and it can be concluded from the results that the proposed algorithm can effectively improve the system throughput.