Network Selection Algorithm Based on Chi-Square Distance in Heterogeneous Wireless Networks

This paper proposes an access network selection algorithm based on a cooperative strategy with Chi square distance, subjective weights for traffic classes and objective attributes of network. The algorithm gets the subjective weights of attribute in four traffic classes using AHP at first, then calculates objective weights using a cooperative method of entropy weight and criteria importance though intercriteria correlation. The final weights are obtained by a comprehensive weighting method based on minimizing the deviation between weights. Finally we measure the Chi square distance between every candidate network and the ideal network and choose the one which has the minimum Chi square distance to the idea network as the best access network. Simulation results show that the proposed algorithm can select the most suitable access network to meet the requirements of current traffic and observably extend the discernibility between candidate networks. The ratio between the best and the second-best network in this algorithm is 8 or 9 times bigger than that in technique for order preference by similarity to ideal solutions, simple additive weighting and grey relational analysis. Furthermore, it can avoid falling into local optimal situation, choose the correct optimal access network when some network attributes appear extreme data, and has good algorithm robustness.