S-GRA And Radar Chart Based Access Selection Algorithm In Heterogeneous Networks

The heterogeneity of wireless networks and the diversity of services have poised a higher demand for radio resource management in the environment of heterogeneous wireless network. Access selection, which is indispensible to radio resource management, is the crucial element affecting the system resource utilization rate. Thus, the design of optimize access selection strategy to guarantee users always be connected to the most suitable accessible network and use the wireless resources of the whole system effectively is a very challenge research project.In this paper, the basic principles and characteristics of access selection in heterogeneous wireless network as well as the advantages and disadvantages of existing algorithms are studied and analyzed, optimize methods are discussed. To reduce the average call blocking probability and balance the network load, Gray Relative Analysis theory and genetic theory and so on are introduced, many factors reflecting users’mobile characters and network performance, such as mobile rate of users, received signal strength, available bandwidth rate and call blocking probability, are overall considered in this paper, which overcomes the single-criterion disadvantage of traditional algorithms and makes the access selection decision more reasonable.Aiming at working out the problem that traditional Gray Relative Analysis control algorithm didn’t give prominent consideration to the load state in heterogeneous wireless network, a S-GRA(Strengthen-Gray Relative Analysis) based on load balance access selection algorithm was proposed. The algorithm took load state as a strengthen factor to adjust resolution coefficient of GRA(Gray Relative Analysis), which can achieve effective load balance as well as guarantee required QoS(quality of service) in each network. The simulation results show that compared with MLB(Maximum Load Balancing) the algorithm could effectively balance the load of heterogeneous wireless networks, guarantee the QoS, decreased and balanced the blocking probability. Meanwhile, compared with traditional GRA which directly took load as one of its factors, the algorithm proposed could obviously balance load when guarantee low blocking probability.Besides, to deal with the problem that the judging factors of access selection algorithm were one-side in heterogeneous wireless networks, a RAD(Radar chart, RAD) based access selection algorithm was proposed. Firstly, the evaluation index matrix should be normalized in order to guarantee equity of each evaluation index when make decision. Then use genetic algorithm, take d-value of evaluation function as fitness function to find the optimal weight allocation scheme. Finally, sort the evaluations that calculate by radar chart algorithm.The maximum value network was the one that will access.The simulation results show that the algorithm can decreased the blocking probability effectively so that assurance the quality of service, meanwhile loads among the overall heterogeneous system can be effectively balanced.