Optimal task allocation and payment minimization strategy of the multihomed end system

Abstract

MEU (multihomed end user) refers to some terminal hosts of enterprises and large data centers that transmit packets to the destination through several paralleled upstream ISPs. In this way the host can improve the performance during the transmission. Currently there are many studies that focused on how to design some optimal and effective strategies to improve the overall performance of the transmission network. To be specific, most of them only aim at finding some balanced allocation algorithms to allocate the task properly. In this paper, we mainly propose a novel task allocation strategy which can be used in a special network topology. We consider multiple multihomed end users which connect to several upstream ISPs. It forms in a many-to-many mapping relation. By formulating the task allocation problems in this scenario, we make use of the game theoretic approach to analyze the issue and achieve the Nash equilibrium in the optimal solution. Next we make the study of the cost consumed by MEUs. By using our task allocation strategy, we calculate the total costs of the multihomed end system. Then we analyze the optimal amount of ISPs and compare our strategy with the equal division algorithm. The result shows that our strategy is better.