An improved sensor deployment scheme for multiple target localization using compressive sensing

In this paper, we propose an approach for multiple target localization in wireless sensor network by employing the compressive sensing (CS) theory, which provides a novel framework for recovering the signal with far fewer sampling values than traditional methods, under the assumption that the signal is sparse. The sparsity in our localization approach is reflected by the location of targets, which can be formulated as a sparse vector. We use the received signal strength (RSS) to achieve the target localization and it only requires a small number of measurements for accurately recovering the location vector by solving the ℓ₁-minimization program. Moreover, we propose an improved sensor deployment scheme, which shows a more stable and valid performance than the random deployment widely used in CS-based localization. Meanwhile, it also decreases the workload of acquiring the sensor locations in practice. Finally, we conduct simulations and the results demonstrate the effectiveness of our localization approach.