Predict the Multi-hop Reliability for Receiver-Contention Based Routing in Dynamic Link Networks
Yongcai Wang, Yuexuan Wang and Dazhong Zheng
We consider dynamic link networks (DLN), where nodes are static, but the links are highly dynamic. Reliable multi-hop data transmission is critical in such networks. Receiver contention routing, which elects relay from the actual receivers are generally proposed to overcome the link dynamics. However, predicting the multi-hop reliability is still difficult, caused by the randomness of packet loss and the packet redundancy in the multihop forwarding process. In this paper, a Markov Chain model is proposed to describe the multi-hop forwarding process, in which the packet loss and the packet reception by the sink are the two absorbing status of the Markov Chain. The multi-hop reliability is predicted by analyzing the stable status of the Markov Chain. Particularly, when the positions of all the sensors are known, the multi-hop reliability of RECORD has closeform formula; when node locations are unknown, recursive upper bound and lower bound are derived. Further, RECORD is implemented in a fully distributed manner in both simulation and real sensor networks. Extensive performance evaluations validate the derived multi-hop reliability results.
Keywords: Dynamic link networks, multi-hop reliability, receiver contention based routing, prediction.