Underwater Wireless Sensor Networks: Efficient Localization Schemes using SemiDefinite Programming

Abstract — Location awareness and distance estimation cannot be underestimated in wireless sensor networks especially for the marine environment. State-of-the-art localization research and approaches have been thoroughly reviewed to illustrate the challenges imposed by inaccurate distance measurements and noisy backgrounds. Compared to existing localization approaches, the SemiDefinite Programming approach delivers accurate distance measurements even in hostile backgrounds. In this paper, two localization schemes, namely Fixed-Position, and Magnified-Range, are proposed. The main idea behind the Fixed-Position scheme is to improve connectivity of the whole system. Magnified-Range is mainly based on the fact that anchors are special nodes that suffer less from energy constraints. Therefore, it is possible to magnify their radio range without impacting the energy consumption of the network. Magnified-Range differentiates between the radio range of anchors and regular sensor nodes. Performance evaluation and simulation results have shown that the proposed schemes offer robust localization. In fact, the Magnified-Range scheme improves localization accuracy by 20%. Future research will focus on three main challenges, namely 3-D simulations, hybrid localization systems, and real-life demonstrations. Keywords-marine wireless sensor networks; underwate