Research On Access Selection Strategies In Heterogeneous Multi-networks Coexistence Scenarios

The wireless technology progress promoted the development and application of global communication industry, also changed the way that people live and work. On the other hand, the increasing communication demands also promote the further development of communication technology. The new communication systems and the technical specifications emerge endlessly. In order to provide users different services and the always-best-connection, in the future, a variety of networks defined by different specifications form Heterogeneous Wireless Networks, instead of networks with one specification.Various wide coverage cellular mobile communications supporting user mobility and short coverage and economy wireless local access networks lead to heterogeneous wireless network coexistence environment, in which different communication systems can provide various services. Users access the best network according to their needs. The next generation network architecture will become more complex and converged, it is inevitable that many heterogeneous networks are highly integrated. In this paper, we studied some access selection strategies in heterogeneous multi-network coexistence scenarios as follows:(1) The research of network selection algorithm of user side: We assume that users get in overlapping area of multiple networks with Poisson probability in a certain time period. The users want maximize their own transmission experience through proper network selection. We proposed an algorithm based on genetic algorithm for single user network selection. The competition between users and networks is not considered in the algorithm. The algorithm uses the sort rule of TOPSIS(Technique for Order Preference by Similarity to an Ideal Solution) algorithm, and dynamically calculated weights and the best network matching rate. It respects users' subjective preference, as well as changes weights dynamically. Users access the best networks which satisfy their own demands in time. We compare the algorithm with AHP+TOPSIS algorithm and AHP+Entropy+TOPSIS algorithm in terms of algorithm convergence, load balance, and network congestion. The simulation verifies the effectiveness of the proposed algorithm.(2) The research of network selection algorithm used by multiple users located in the same area and apply the same service, for example in bus stations, shopping malls, and large CBDs. We model the network selection problem for multiple users select network in heterogonous wireless network scenarios as a game, then propose an algorithm based on evolution game. The proposed algorithm is compared with RSSI(Received Signal Strength Indication) algorithm in terms of average users' benefits, resource utilization.