Research Of Interference Management Techniques In Ultra Dense Networks

With the popularization of smart devices and the rapid development of mobile Internet,the demand of the transmission rate of wireless communication is increasing.In the fifth generation mobile communication system,the ultra-dense networks is one of the key technologies that can effectively improve networks throughput and spectrum efficiency.The densely deployed nodes reduce the distance between the user and the access point greatly.While increasing the user rate at the same time,it also brings about the problem of interference inevitably.This thesis starts with the analysis of interference modeling in ultra-dense networks and studies the problem of interference management in ultra-dense networks.At present,there are existing interference management technologies for traditional cellular networks,including interference cancellation,interference avoidance,and space diversity.However,due to the heterogeneity and complexity of deployment of ultra-dense networks nodes,it is necessary to re-evaluate the effectiveness of these interference management methods.At the same time,how to apply machine learning methods to build intelligent interference management systems is also a open question worth studying.In order to solve the interference problem caused by the densely deployment of networks,this thesis starts from the modeling and analysis of interference,leveraging stochastic geometry to model the location of base stations and users,and studies the problem of joint cluster transmission and inter-cell interference coordination in ultradense networks.Lastly,reinforcement learning is also investigated to optimize the parameters of inter-cell interference coordination.The main contribution of this thesis is described as below:Firstly,the system performance while using the joint cluster transmission mode in the ultra-dense networks is studied.Through the joint transmission of several base stations to users,the dominant interferers can be transformed into useful signals,thereby greatly increasing the user rate.The Poisson point process is used to model the location of the ultra-dense networks base station and the user,and the active probability of the pico base station is given.The expressions of the coverage probability and the downlink average rate are obtained.The effect of the pico base station density and the cluster size on the system performance are analyzed.The numerical analysis and simulation results reveal the relationship between the active probability and the microstation/user ratio in the ultradense networks.Increasing the size of the joint transmission cluster can increase the user rate.The increase in the deployment density of microstations can also increase the total networks rate,but this gain is diminishing when the pico-base station density grows large.Then,towards the problem of inter-cell interference coordination in ultra-dense heterogeneous networks,the user’s average spectrum efficiency in this case is analyzed using stochastic geometry modeling.The influence of cell selection bias and power control parameters on system performance is analyzed.The numerical analysis and simulation results show that the average spectrum efficiency of the user can be maximized when using reasonable inter-cell interference coordination parameters.Finally,in order to select the optimal inter-cell interference coordination parameters,a behavior-based learning algorithm for inter-cell interference coordination is proposed.The algorithm enables the macro base station and the pico base station to independently select the optimal cell extension bias and power control parameters,which maximizes the system capacity.