Research On Access Selection For Heterogeneous Wireless Networks Oriented To Smart Distribution Grid

The development and safe operation of smart grid cannot be separated from the support of information and communication technology.Broadband wireless communication network has the advantages of flexible access,strong business support ability and low investment cost It has become a recognized technology trend in the international power industry.However,the current single access technology cannot fully meet the diverse requirements of distribution network communications.Thus,it is one of the effective solutions to adopt multiple access technologies to complement each other and form heterogeneous access networks.In order to ensure the timeliness and reliability of the transmission of distribution communication services based on the analysis of the demand of distribution communication services,heterogeneous wireless network architecture for smart distribution network is proposed in this thesis.Under the architecture,communication terminals are multi-mode devices which can access different networks to obtain services to meet their own QoS requirements.How to choose the network becomes a key issue of resource management in heterogeneous wireless network.At the same time,how to choose the network becomes a key technology of resource management in heterogeneous network environment.Based on the network access selection of communication terminals in heterogeneous wireless network architecture,this thesis analyses the issues that need to be paid attention to network access selection.To ensure the stability of QoE and power system,on the basis of traditional QoE(quality of experience)metric,delay factor is introduced to weight delay's effect on service transmitting,and both delay and rate requirements are considered to measure the QoE.Firstly,the service delay and rate utility functions are defined respectively,and the service delay factor is derived from the delay utility function,and the QoE metric based on the delay factor and rate utility is constructed.Then,the resource management scheme is modeled as an optimization problem,which is a mixed integer nonlinear programming(MINLP)problem.Genetic algorithm(GA)is utilized to solve it.The simulation results show that the proposed access selection algorithm can satisfy the user's delay requirements and guarantee the user's rate utility,and improve the load balancing performance of the base station.Due to the limited power of communication terminals' batteries,a low power access selection algorithm is proposed to prolong the working time of communication terminals in distribution grid.Considering the user's service quality requirement and the residual power of auxiliary power supply in distribution communication terminals,an access selection scheme based on weighted power minimization is constructed.Mathematically,the model is MINLP problem.Since each user accesses at most one base station,the transmit power matrix from user to base station is a sparse matrix,and the MINLP problem can be transformed into a sparse solution to the optimization problem.In this paper,the L norm is introduced as a penalty function to replace the integer constraint and to constrain the sparsity of the transmit power matrix from the user to the base station.And a resource management algorithm based on reweighted L norm penalty is designed.The simulation results show that the resource management algorithm based on reweighted L1 norm penalty has good convergence,and the results are basically the same as those of genetic algorithm.The proposed access selection algorithm can improve the online rate of communication terminals and prolong the working time of low-energy users.