Research On Local Delay And Energy Efficiency In Heterogeneous Networks Via Stochastic Geometry

With the exponential growth of wireless data traffic,mobile communications technologies and industries have entered the development stage of the fifth generation systems(5G).Compared with the fourth generation mobile communications systems,5G requires the mobile communications system to have much higher service throughput,massive load connections,1000 times improvement of spectrum efficiency,10 times energy consumption reduction and millisecond latency.Heterogeneous networks are one of the important candidate technologie,which can potentially meet the requirements of 5G for latency and energy efficiency.However,the commonly used distribution model of base stations of heterogeneous networks is usually the Poisson point process(PPP),ignoring the actual spatial coupling characteristics.In addition,latency and energy efficiency are important performance metrics for measuring the reliability and effectiveness of heterogeneous networks.Therefore,this thesis will study the latency and energy efficiency of heterogeneous networks via stochastic geometry from the perspective of base stations and users.The term "latency" here mainly refers to the transmission delay,also called the local delay.Aiming at the problems of existing distribution models of base stations,this thesis uses the Poisson cluster process(PCP)to model base stations,and uses the discontinuous transmission method to manage the interference of base stations and reduce energy consumption.Based on the heterogeneous networks of the PCP,the expressions of local delay and energy efficiency are derived.Local delay and energy efficiency with respect to key system parameters,such as the mute probability,base stations' density,are analyzed.At the same time,the differences on the local delay and energy efficiency under PPP and PCP are compared.The simulation results show that the performances of the two models are similar under the condition of low signal-tointerference ratio threshold or high mute probability.The mobility of the user reduces the correlation of the base stations' interference,further affecting the local delay and energy efficiency.The asymptotic analysis method is adopted—assuming that the user's speed and move scope are infinite,and the user's infinite mobility is equivalent to the base stations' movement.Based on such a heterogeneous network model,we first derived the corresponding local delay and energy efficiency,and then analyzed the impact of user mobility on the local delay and energy efficiency as well as the performance of discontinuous transmission in a dynamic heterogeneous network.The simulation results show that the user's mobility will reduce the performance of the discontinuous transmission method,but will not affect the role of the cluster.Finally,considering that the low-power base stations tends to have small loads and the macro cells are likely to have large loads in the heterogeneous networks,this thesis uses the biased association to extend the coverage of low-power base stations and solve the load balancing problem in heterogeneous networks.By introducing a bias factor to the user's received signal,the appropriate bias factor is established with the per-user bandwidth,and the local delay and energy efficiency of the biased heterogeneous networks are derived.Then,the bias factor is optimized to improve the local delay and energy efficiency of the heterogeneous networks.Simulation results show that the optimal biased association can significantly improve the performances of local delay and energy efficiency.