Network-Assisted Sink Navigation Protocols for Data Harvesting in Sensor Networks

Abstract

This paper presents a network-assisted sink navigation mechanism in which mobile sink trajectories are computed by a sensor network for energy-efficient data collection. The key idea is that for data collection, a mobile sink (referred to as Mobile Data Harvester MDH) should visit all network nodes in order to avoid energy- intensive multi-hop routing. While incurring higher data collection delays compared to the traditional static sink based collection, the proposed harvesting framework can deliver the maximum achievable lifetime for delay non-critical sensor applications. The main issue in this approach is to compute MDH trajectories that can reduce the average data collection delay while traversing through the radio ranges of all nodes, but not necessarily pass through the close proximity of each node. Our primary contribution is the design of a fully distributed and network assisted MDH navigation mechanism where MDH trajectories are cooperatively computed by the nodes without sensor localization services. Performance results obtained from an ns2 simulator indicate that the distributed algorithms and network assisted navigation framework can achieve low average delay, low delay variation and complete collection coverage of the network. Further, the performance of the distributed algorithms was found to be comparable to that of centrally computed solution.