The Research On Service Access Control Method For HWNs Based On Multi-objective SOS Algorithm

In order to solve the problem of service access control in Heterogeneous Wireless Networks(HWNs)and improve the utilization of network resources while ensuring the quality of service to users,this paper constructs the mathematical model by the performance indicators that have significant impacts on the access control of HWNs,proposes a new evolutionary algorithm with excellent performance to optimize the model and achieves the best access control schemes.The main research contents include two aspects: one is proposing a new multi-objective optimization algorithm,the other is solving the model of service access control in HWNs by the proposed algorithm and verifying the optimization effect of the algorithm in practical application.Besides,the proposed algorithm is the combination and improvement of the Multi-objective Evolutionary Algorithm Based on Decomposition(MOEA/D)and Symbiotic Organisms Search algorithm(SOS),multiple performance indicators that affect service access control are optimized by the proposed algorithm simultaneously.The research contents are introduced as follows.First of all,SOS is taken as the core evolutionary strategy of the whole multi-objective service access control algorithm and its performance has important impacts on the optimization results of service access control.In order to improve the optimization capability of SOS,this paper proposes an Elite Symbiotic Organisms Search algorithm based on subpopulation stretching operation(SPS-SOS).In mutualism phase,the population is divided into two subpopulations according to the fitness value,different subpopulations adopt different evolutionary methods,which keeps a good balance between the development and exploration.In Commensalism phase,introduce the stretching factor and the difference perturbation vector and modify the individual updating mode,which improves the convergence speed and maintains the population diversity.Besides,the parasitism mechanism of the elite further balances the development and exploration of the algorithm.The experimental results show that SPS-SOS has obvious advantages compared with other evolutionary algorithms.Secondly,in order to improve optimization ability of evolutionary algorithms in multi-objective problems,this paper proposes a multi-objective Symbiotic Organisms Search algorithm based on Decomposition(MOEA/D-SOS)which combines MOEA/D and SPS-SOS.To improve the convergence of solution set,take SPS-SOS as the core evolutionary algorithm of multi-objective optimization algorithm,modify the individual updating and retaining methods according to the individual evolutionary characteristics of MOEA/D,which makes the SPS-SOS more suitable for the updating way of multi-objective optimization algorithm.To improve the search efficiency of the MOEA/D-SOS,adjust the neighborhood sets dynamically during evolution.To improve the distribution of solution set,regenerate the weight vectors by the characteristics of the solutions themselves in later evolutionary stage.The experimental results show that MOEA/D-SOS has significant effect on multi-objective optimization problems.Lastly,in order to save network resources and guarantee the quality of service at the same time,MOEA/D-SOS is used as the core control algorithm to solve service access control of HWNs.Establish the multi-objective mathematical model which is aimed at minimizing business occupancy,minimizing network congestion rate and balancing the network load,solve the mathematical model by MOEA/D-SOS directly,the retained non-dominated solutions after non-dominated sorting are the optimal access control schemes.Compared with other access control algorithms,the experimental results show that MOEA/D-SOS has better performances on improving the utilization of network resources,reducing network congestion rate and balancing the network load.