Investigations On The Wireless Network Resource Allocation Based On User Behavior Analysis

With the development of Internet technology and the popularization of smart phones,various mobile phone applications have greatly enriched people's lives.At the same time,people's demand for mobile network traffic has also shown explosive growth.In order to meet people's increasing communication needs,mobile operators use ultra-dense networks,massive MIMO and other technologies to increase system capacity and provide users with better mobile communication services.However,the addition of hardware equipment such as base stations and antennas has increased network capacity,and on the other hand,has significantly increased the overall power consumption of mobile networks.Therefore,how to reduce the energy consumption of the network while meeting the needs of users has become the focus of people's attention.In this paper,we propose a base station antenna dynamic switching method based on user behavior prediction in large-scale MIMO systems.The method is based on principal component analysis algorithm and clustering algorithm to cluster users.Then,based on wavelet transform and LSTM network,we predict the user's traffic consumption and estimate the user's required Qo S.The number of base station antennas is dynamically adjusted based on the user's Qo S prediction results,so that the number of base station antennas turned on can provide users with sufficient Qo S to achieve the goal of reducing network energy consumption in the case of meeting user needs.The simulation results show that the method of adjusting the number of base station antennas based on user traffic prediction compared to the method based on the number of users can significantly reduce system energy consumption at the same time under the same user satisfaction rate.For ultra-dense networks,we propose a small cell wake-up strategy based on user behavior prediction.This method is based on the LSTM network to model the user's APP usage behavior,predict the type of APP the user will use in the next period and estimate the Qo S required by the user.The macro base station judges whether the small base station needs to be awakened according to the user's Qo S prediction result.When the user's Qo S requirements are low and the macro base station can meet the user's needs,it does not need to wake up the small base station,and only needs to be awakened when the macro base station cannot meet the user's needs.Simulation results show that compared with the method based on the number of users,the small cell wake-up strategy based on user APP usage behavior prediction can significantly save system energy consumption under the same user satisfaction rate.