Abstract:

Device-free localization (DFL) is the estimation of target without carrying any electronic device by the shadow-fading of wireless communication links. However, current algorithms have some disadvantages of low positioning accuracy and large computation during positioning. A novel algorithm based on radio tomographic imaging(RTI), combining regularization and compressive sensing theory, called bi-reconstruction algorithm, is proposed to locate the target. The algorithm uses the characteristics of the fast reconstruction of regularization and first of all, it divides the location area into a plurality of grids to complete the initial positioning of the target. In order to achieve precise positioning, the possible location areas are divided into multiple small grids again, simultaneously the link selection method is used to decrease the number of links and reduce the algorithm complexity. Then, complementary sparsity adaptive matching pursuit (CMPs-SAMP) is adopted to locate the target and transform the localization problem to a sparse signal reconstruction problem. The simulation results show that the proposed bireconstruction algorithm outperforms other single reconstruction algorithms.