Traffic distribution method for concurrent multipathtransfer based on maximum revenue
多通道并行传输中基于收益最大化的流量分配方法

In order to solve the problem that existing traffic distribution methods for concurrent multipath transfer do not distinguish services or consider the revenue of resource, this paper analyzed service parameters and used the revenue of services as the benefit evaluation function of the traffic distribution for concurrent multipath transfer. It built a traffic distribution model based on maximum revenue through mapping traffic distribution to bandwidth allocation. Based on the law of diminishing marginal utility, it proposed a traffic distribution method with marginal revenue priority for concurrent multipath transfer. The algorithm firstly obtained the traffic distribution results which had maximum revenue by a greedy strategy with marginal revenue priority, and then adjusted the traffic among the paths to ensure the minimum QoS. The case analysis and experimental results verify the effectiveness of the method. Compared with the traffic distribution method based on bandwidth ratio policy, the proposed method can achieve the optimal allocation of bandwidth resource and improve the operator's revenue while ensuring the service QoS.