Data Gathering in Opportunistic Wireless Sensor Networks

The wireless sensor networks and opportunistic networks have nowadays presented a trend of technology convergence. On one hand, the nodes periodically sense the environment and continuously generate sensing data; on the other hand, the movements and sparse deployment of nodes usually lead to intermitted connected links and create some form of opportunistic communications. So it is a challenging problem to effectively collect the sensing data in opportunistic wireless sensor networks. In this paper, we propose an efficient data gathering algorithm based on location prediction in opportunistic wireless sensor networks. The algorithm first collects the network metadata such as history of node encounters and contact durations; then it creates a node contact graph, based on which predictive optimal data gathering locations are dynamically calculated and updated. Finally, the sink is controlled to move to these locations to collect sensing data, avoiding lots of unnecessary data exchanges and message transmissions. Extensive experimental results show that the proposed algorithm is effective to reduce the message transmissions and improve the data collection coverage rate.