Multiple Coverage with Controlled Connectivity in Wireless Sensor Networks
Résumé
The k-coverage of the area of interest is a classical requirement in wireless sensor networks, for purposes such as redundancy or multiple supervision of the area. We aim in this article at preserving full k-coverage of the area of interest yet minimizing the number of sensors needed for data gathering. Our main contribution consists in a localized layer-based coverage protocol and introduces a "Sensing-Only" state for some of the monitoring sensors. Indeed, among active sensors that achieve multiple coverage, a small connected subset can be constructed in order to ensure data multi-hop transmissions from points of interest to sink stations, thus allowing remaining sensors to cut their radio in reception. We evaluate our scheme through different constructions of sets of eligible nodes, ranging from a simple layered k-coverage to a more elaborated cross-layer determination. Ensuring multiple coverage with a reduced connected active node set, the present solution does not only reduce the overall activity of the network and the number of emissions, but also requires only few extra messages to be efficient.