| Self-Organizing Network(SON)is a set of network concepts and standards developing along with the Long Term Evolution system.The design concept of SON is to realize the autonomous functions of wireless network,mainly including self-configuration,self-optimization and self-healing which is suitable for the Heterogeneous Ultra-Dense Network(H-UDN).Under the deployment of H-UDN,the increasing demands for data traffic and the more and more complex network deployment structure bring great challenges to network performance optimization.On the one hand,denser deployment of network leads to more frequent handovers and higher handover failure rate.On the other hand,the configuration of handover requires dynamic adjustment due to the multiple communication services.At the same time,H-UDN is faced up with the urgent needs for connection of massi-ve terminal equipment as well as the shortage of spectrum resources.Thus,the mechanism of Device-to-Device(D2D)communication requires further improvement.Therefore,aiming at the problems above,we carried out the research work of this thesis:Focusing on the H-UDN deployment,we studied the mobility management strategy.As the important part of mobility management,handover process is divided into four key modules,including wireless link monitoring,measurement reporting,handover decision and handover execution.In addition,Stochastic Geometry,an important analysis method in H-UDN,is a key modeling tool for handover.The detailed analysis of the Poisson Point Process and Random Waypoint Movement Model involved in handover process lays the foundation for the following works.Aiming at reducing the high rate of cross-tier handover failure in H-UDN,this thesis proposed a handover parameter self-optimization strategy based on analytical results applying Stochastic Geometry.Firstly,the closed-form equations of handover failure rate and ping-pong rate are derived using Stochastic Geometry.Under the guidance of the equations,we made a prediction of handover parameters.Then,on the basis of the traditional handover strategy,considering the influence of the Reception Signal Strength,Sojourn Time and Traffic Load on handover performance,combining the fuzzy comprehensive analysis method,the weights of the three parameters are calculated.Finally,with the weights of the three parameters as input,the grey correlation degree analysis method is adopted to calculate the correlation degree between the candidate base station and the optimal base station.System-level simulation shows that the handover strategy proposed in this thesis effectively improves the handover performance by reducing the handover failure probability and ping-pong probability,as well as improving the data rate.In terms of H-UND D2D communication,the self-allocation strategy of D2D discovery resources is proposed with the goal of enhancing the discovery process with higher efficiency and lower energy consumption.In current procedures,the arrangement of announcing and monitoring resource pool doesn't take their potential relations into consideration.Thus,in the view of actual D2D communication scenario,we optimized the discovery procedure based on the historical connection information.Firstly,the Relay UE and Remote UE historical Connection Times are recorded through the Advanced Trust Relationship List.After the base station processing,the priority lists of Relay UE and Remote UE are obtained.Therewith the resource pool is allocated according to the priority lists in advance,and the self-configuration of discovery resources is realized.Simulation results show that the self-allocation strategy proposed in this thesis can effectively reduce energy consumption and improve discovery efficiency.To sum up,in this thesis,the strategies applying self-optimization and self-configuration are carried for cross-tier handover and D2D discovery in H-UDN.Through problem study,model building,process design and simulation analysis,it is proved that the proposed strategies can achieve theoretical performance gains. |
PDF Full Text Request