期刊名称：International Journal of Distributed Sensor Networks
印刷版ISSN：1550-1329
电子版ISSN：1550-1477
出版年度：2011
卷号：2011
DOI：10.1155/2011/939864
出版社：Hindawi Publishing Corporation
摘要：In Underwater Acoustic Sensor Networks (UASNs), data route is often disrupted by link interruption which will further lead to incorrect data transmission due to high propagation delay, Doppler effect, and the vulnerability of water environment in acoustic channel. So how to correctly transmit data when there are interrupted links on the data path is just the issue Delay Tolerant Networks (DTNs) aim to solve. In this paper, we propose a model to predict link interruption and route interruption in UASNs by the historical link information and channel state obtained by periodic detection. A method of decomposing and recomposing routes hop by hop in order to optimize route reselection is also presented. Moreover, we present a back-up route maintenance scheme to keep back-up routes with fresh information. In case of single route, we advance the idea to utilize the periodicity of environmental changes to help predict link interruption. In the simulation, we make comparisons on node energy consumption, end-to-end delay as well as bit error rate with and without link prediction. It can be derived that the network performance is significantly improved with our mechanism, so that our mechanism is effective and efficient while guaranteeing reliable data transmission.
