Research Of Resource Sharing Mechanism In Mobile Social Networks

DTN (Delay Tolerant Network) is a kind of challenged network usually deployed in extreme areas. Buffer size, wireless link and contact opportunities may become the bottle neck of the network. These environment and characteristics make it difficult to communicate and sync between nodes. Due to the intermittently connection and sparse distribution, routing in single copy manner performs badly. Many researches propose multi-copy routing methods. But multi-copy routing increases system cost largely. In case the network load gets high, the over consume of resource will cause many problems. For example, the over consumption of buffer will cost congestion; over consumption of energy will cause nodes dying too soon. Compared to traditional networks, DTN lacks researches about resource management.The unreasonable allocation of resources will make bad effects on the network performance. In DTN using social routing, traffic is converged on nodes with higher social metrics. When there is a heavy load, such nodes are facing higher possibility to be congestion. Many researchers have proposed methods to alleviate this congestion. However, congestion is in fact due to the over consume of resource in nodes. It is more appropriate to solve this problem from a resource allocation aspect.Fairness is an important indicator of network performance. In actual social DTN, individuals want to get at least the same network experience with others even when the network resource is tight. This paper takes fairness as a goal of resource allocation. At the same time, available resources should be full exploited to serve network users. So network utility is taken as the other optimization goal.This paper estimated the level of users’resource occupation by the replica quota of multi-copy routing. Then, I analyzed the lack of fairness in the allocation of resources in current DTN. The relative stability of human activity is discovered and evaluated after the regularity of human activity is analyzed. Furthermore, though such stability and characteristics of social routing, I found path stability of social routing. Based on this stability, a mechanism to control resource allocation of DTN is proposed. This mechanism can ensure the fairness of resource allocation under certain conditions. At the same time, network utility can be maximized conforming to resource limits.