Resource Optimization Based On User Behavior In Heterogeneous Networks

Recently,wireless communication technology and Internet technology have developed rapidly,wireless services such as voice,video,data services creates an overwhelming demand on data traffic,the heterogeneous networks(Het Nets)comes into being.In Het Nets,the coexistence of Macro base station and low-power base station can improve the performance of cell coverage and capacity and meet the needs of high-speed and multi-service of the users at the same time.However,as the users' demand for higher throughput and greater peak traffic grows,energy consumption in Het Nets is increasing,energy efficiency has become a principal metric of the communication system's performance.Meanwhile,the limited resources in the communication network have to be used and allocated rationally.Therefore,the problem of resource allocation in Het Nets has become a research focus.With the emergence of various wireless services,users tend to get all kinds of information from the network,then increased user experience,the user's personalized demand arise.The analysis of user's network behavior has become a vital issue in the field of mobile communications.Currently,few of the research study the behavior of mobile users,on the basis of which to do the resource allocation in Het Nets.This paper investigates the resource optimization problem based on user behavior and the improvement of system energy efficiency in Het Nets.In this paper,the behavior of the mobile user is first investigated.Based on the D-PER(Density-Preferential Exploration and Return)individual movement model,the basic movement pattern such as radius of gyration,frequently visited locations is analyzed.Integrating with the periodicity of the user's movement,the next location that the user is going to visit,the requested service type and requirement for Qo S in the next location are predicted.Finally,the user behavior prediction algorithm based on D-PER model is developed.Compared with other algorithms,the proposed algorithm can achieve higher accuracy for the prediction of the next location,and the requested service type and Qo S requirement in the next location.Further,the distribution of downlink wireless resources based on user behavior in Macro-Femtocell Het Nets is studied.And the active array 3D beamforming technology is introduced into the existing network while considering the mobile user behavior,such as location,service type,Qo S requirement.Some users are clustered into several user groups according to the location and request preferences,which are served by multicast transmission aided by the active antenna beam.While users outside the clusters are served in a traditional manner.The Dinkelbach method and the Lagrangian dual decomposition method are used to solve the resource optimization problem to maximize the energy efficiency.A joint user association and resource allocation algorithm is devised which can improve the resource utilization while ensuring the Qo S requirement of users.Beside the optimization of the energy efficiency of the base station in Het Nets,considering terminals' energy consumption problem and the users' statistical delay constraint,the uplink resource allocation problem is investigated which minimizes the terminal power consumption.In this scenario,the concept of effective bandwidth(EB)and effective capacity(EC)is introduced to describe statistical delay constraints.The problem of minimizing the terminal power consumption is formulated as a convex optimization problem which can be solved by KKT condition.An iterative resource allocation algorithm is designed to reduce the power consumption and meets the principle of green communications under the premise of satisfying the statistical delay constraint.