Abstract
Data exchange is one of the main reasons that social network takes form. In application of ad hoc wireless network, appropriate information sharing for demanders helps reduce the communication overhead. Thus, social routing is a new research direction. However, we find, currently most social networks work in a centralized way, they always collect other’s information in their environment and actively predict utilization for their route. And we also find there are less social routing researches related to on-demand routing protocols, because they are not proactive for routing.Actually, these on-demand routing protocols are aimed to save more energy and consume less bandwidth; their features make them more and more popular recently. In this paper, we introduce the concept of using tag matching for routing information prediction and intermediates nodes can savethe transfer data by pre-enumerated person list matching with tags. We implement simulation among on-demand routingprotocol AODV, and evaluation its performance. We except to keep the advantage of AODV, simultaneously enhance the utilization of forwarding data that would cost down communication overhead.