Stochastic Geometry Theory Based Research On Dense Heterogeneous Networks

Modeling and performance evaluation of cellular networks is very important in the field of wireless communication.Traditional grid modeling cellular networks are often not analytical and require complex system level simulation.With the development of future heterogeneous networks,network nodes show more random and denser distribution,which increases the complexity of traditional system level simulation.In recent years,stochastic geometric theory has been applied to the modeling and performance analysis of multi-tier dense heterogeneous networks.The stochastic geometric model makes up the defects of the traditional system level simulation,and can describe the randomness of the node deployment of the multi-tier dense heterogeneous network,and can get a lot of accurate analysis results of the network performance.Therefore,stochastic geometry based modeling and analysis of multi-tier dense heterogeneous cellular networks has attracted more and more attention.Against this background,this thesis choose to use stochastic geometric theory to model and analyze dense heterogeneous networks.Our work solves the problems of mobility management and network architecture evolution in the previously unresolved dense heterogeneous networks,and the relevant work provides reference for other researchers.Our work is to analyze the mobility management and network architecture of the multi-tier dense heterogeneous network by using the stochastic geometry theory and mathematical tools such as probability theory and convex optimization theory.Along with this line,we have done important handover metrics analysis,handover optimization and network architecture evolution of multi-tier heterogeneous networks analysis.Specifically,this thesis covers three aspects:1.The cross-tier handover process in heterogeneous networks is analyzed based on stochastic geometry theory.Based on the 3rd Generation Project protocol,a closed-form expression of small cell boundary and handover failure boundary is obtained through analysis.According to the proposed small cell and handover failure boundary model,the expression of three important handover performance metrics including handover rate,handover failure rate and ping-pong rate are derived.The closed-form solution shows the handover rate,handover failure rate and ping-pong rate can are influenced by Time-to-Trigger value,the density of base station deployment,and the velocity of users.The system simulation results verify the correctness of the theoretical derivation.These results provide theoretical guidance for the network deployment and the optimization of the mobile strategy in dense heterogeneous networks.2.The performance of cross-tier handover in heterogeneous networks is optimized based on stochastic geometry theory.First,the effective capacity of users before and after cross-tier handover is analyzed,and a simple solution of effective capacity is obtained.Then a new resource reservation scheme is proposed,and the corresponding blocking probability expression is derived based on the resource reservation scheme.Finally,the convex optimization theory is used to optimize the effective capacity and congestion rate.The simulation results can help the operators to select the handover strategy and the resource reservation schemes in actual heterogeneous networks.3.On this basis,the path and method of radio access network evolution in the future mobile networks are studied.Based on the Frameless Network Architecture proposed earlier,the important challenges in the evolution of the cloud wireless access network are analyzed.Then,the evolution scheme and network modeling method of the cloud radio access network based on frameless network architecture are proposed.It can achieve user centered resource allocation and control plane/user plane separation and adaptation.Aiming at multi-objective andmulti-dimensional resource allocation,a centralized management scheme based on genetic algorithm is proposed.It provides a possible direction for the evolution of mobile networks.