TDL: Two-dimensional localization for mobile targets using compressive sensing in wireless sensor networks

Many applications in wireless sensor networks (WSNs) ( e.g. , traffic monitoring, environment surveillance and intruder tracking) rely heavily on the availability and accuracy of targets’ locations. Compressive sensing (CS) has been widely applied to localization as it asserts that a small number of samples will suffice for sparse or compressible signal recovery. Despite much progress in CS-based localization, existing solutions mainly consider static targets and often perform poorly for mobile targets. In this paper, we develop a novel two-dimensional localization (TDL) framework for mobile targets using compressive sensing. TDL is composed of two modules: (i) spatial localization module (SLM) that first achieves localization at sampling times by exploiting the sparse nature of Received Signal Strength (RSS) vector in space domain, and (ii) temporal localization module (TLM) that then achieves localization at all times by exploiting the compressible nature of location vector in time domain. Furthermore, two practicable measurement matrices are constructed to conduct linear measurements. We analyze the flexibility and effectiveness of TDL in theory. Extensive numerical evaluations with real mobility traces further confirm the superior performance of our localization framework. © 2015 Elsevier B.V. All rights reserved.