Heterogeneous Network Selection Algorithm Research

One of the characteristics of the next generation of mobile network is the fusionbetween heterogeneous networks.Different wireless access technologies provide userswith different bandwidth, coverage area, rate support and service quality. To select anappropriate network in the coverage area where contains more than one wireless accessnetwork is one problem that must be solves by the user. In this thesis, we put forwardtwo kinds of network selection algorithms to solve this problem.This thesis puts forward a kind of multi-objective decision network selectionalgorithm. First of all, the parameters which affect the choice of users will benormalized in this network selection algorithm, according to the user’s business, thisthesis calculate the weight of the various parameters through the analytic hierarchyprocess (AHP), the user will select the network that has the largest utility value.Compared to the traditional network selection algorithm based on strength of receivedsignal strength (RSSI), the multi-objective decision network selection algorithm basedon the analytic hierarchy process (AHP) has better performance.This thesis puts forward a kind of network selection algorithm betweenheterogeneous networks based on game theory. Game theory, as a kind of mathematicaltheory to analysis influence between the participants has been widely used in economics,but not widely used in communication network. This kind of network selectionalgorithm is based on matching game, matching game is first presented by Gale andShapley in order to solve the marriage matching problem between men and women. Thenetwork selection algorithm based on the game theory achieves a balance between theusers and networks, then realize the win-win situation of networks and users. First of all,this thesis sets the satisfaction functions for the networks and users, then get the prioritysequence of the users and networks through calculate satisfaction functions of the usersand networks. Finally we can get the matching result through the matching process. Thestability of the game model of the algorithm is proved in this thesis in order to preventthis situation that a better matching stops the original match. Finally, through thesimulation, we can find the performance of the network selection algorithm based on game theory is superior to the network selection algorithm based on user priority andnetwork priority. As the number of users and networks increased, the network selectionalgorithm based on game theory has more obvious advantage.