A Network Selection Algorithm Based On Combined Weights

[摘要] Heterogeneous network selection is a typical multiple attribute decision making problem. This paper studies the determination method of parameters weight in the network selection process and proposes a network selection algorithm based on combined weights. The algorithm respectively used the G1 method and entropy method to determine the subjective weights and objective weights of judgment index, through a linear combination of the subjective and objective weights to get the optimal combination weights. At the same time, make use of network evaluation model based on improved TOPSIS to conduct comprehensive ranking for each candidate network and choose the best one. Simulation results show that the algorithm can obtain lower average latency, jitter, packet loss rate and etc., compared to a typical network selection algorithm based on single weights.

[发布机构] China Satellite Maritime Tracking and Controlling Department, 103# Jiangjin, Jiangsu, China^1

[效力级别] 无线电电子学 [学科分类] 计算机科学（综合）

[关键词] Determination methods;Heterogeneous network selections;Linear combinations;Multiple attribute decision making problems;Network evaluation;Optimal combination weight;Packet loss rates;Subjective weights