Research On Energy Efficiency Optimization Of Heterogeneous Cellular Network Based On Two-Dimensional Distribution Of Traffic In Time And Space

With the continuous development of mobile communication technology,various types of services are emerging in an endless stream,and the size of cellular networks also grow along with the growth of services,leading to the energy consumption problem of cellular networks becoming increasingly prominent.In the traditional deployment of cellular networks for peak traffic,the base station with low traffic or no traffic is still full load,causing a lot of energy waste.Therefore,it is imperative to make full and rational use of communication resources and optimize the efficiency of cellular networks.Network-level energy efficiency optimization plays a crucial role in the energy efficiency of cellular networks.At present,most network-level energy efficiency optimization programs are designed based on the distribution of traffic.Therefore,studying the distribution rules of traffic can help us to better optimize cellular networks efficiency.The traditional research on the distribution rules of traffic is more focused on the single dimension of time or space.At the same time,the research on the two-dimensional spatial and temporal distribution of time and space considering the time and space has drawn wide attention in the industry.Thesis topic comes from the Beijing University of Posts and Telecommunications Youth Science and Technology Innovation Program"multi-group information representation and low-energy content distribution networking"(project number:2014ZD03-01).This thesis focuses on how to optimize the efficiency of cellular networks for traffic distribution based on the combing of energy efficiency optimization and traffic distribution in existing cellular networks.This paper proposes a cellular network traffic distribution prediction algorithm based on Bayesian spatio-temporal model,which can better predict the spatial and temporal distribution of traffic of cellular network.Thus,depicting the two-dimensional spatial and temporal distribution rules of traffic.According to the two-dimensional spatial and temporal distribution of traffic,two cell zooming algorithms for different user densities are proposed in heterogeneous cellular network model to reduce the network energy consumption and improve the network energy efficiency.The main contents of this thesis are as follows:(1)The paper analyzes and sorts out the classification of energy efficiency in cellular networks and the research status of energy efficiency optimization based on cellular network business distribution in detail,as well as the challenges and challenges faced by them.It also points out the potential of optimizing energy efficiency according to the space-time two-dimensional distribution of traffic.This paper summarizes the research status,research contents and challenges of the current cellular network distribution rules of traffic and points out that there is great potential for the current research on the two-dimensional spatial and temporal distribution of traffic.(2)The research on the space-time two-dimensional distribution rules of traffic mainly focuses on the spatial-temporal correlation and modeling,focusing on describing the two-dimensional spatial and temporal distribution of traffic through the induction and theoretical analysis of historical information.The prediction of two-dimensional spatial and temporal distribution of traffic can also help us understand more clearly the two-dimensional spatial and temporal distribution of traffic.Therefore,this paper proposes a cellular network traffic distribution prediction algorithm based on Bayesian spatio-temporal model to predict the spatial and temporal distribution of traffic information.The algorithm considers the influence of the past traffic distribution on the future traffic distribution and corrects the parameters according to the actual data.Simulation results show that the prediction accuracy of different regions can reach above 0.9.And through the forecast results more intuitive analysis of traffic distribution rules.(3)In energy efficiency optimization based on the rules of traffic distribution,cell zooming has great potential due to its flexibility in adapting to traffic needs and optimizing energy efficiency.In view of the spatial and temporal distribution of traffic in the two-dimensional spatial and temporal distribution of traffic,this paper presents LB(Load Balance)cell zooming algorithm and LC(Load Concentration)cell zooming algorithm in heterogeneous cellular network model.The simulation results show that the two algorithms can effectively improve the network energy efficiency under different user densities.The LB cell zooming algorithm is suitable for scenarios with high user densities.The LC cell zooming algorithm is suitable for scenarios with low user densities.In this paper,the spatial-temporal distribution of traffic information is used to clearly understand the two-dimensional spatial and temporal distribution of traffic.According to the distribution rules of traffic,a cell zooming algorithm suitable for different user density is proposed to optimize energy efficiency in heterogeneous cellular network model,which provides a new idea to cellular network efficiency optimization.