Two stages signal strength difference localization algorithm using SDP relaxation

We present a novel two stages signal strength difference (TS-SSD) localization algorithm in this letter. A new model using TS-SSD technique is derived to eliminate the effects of path loss exponent and unknown transmit power. And a total least squares (TLS) solution is given to estimate the distances between anchor and target nodes. Then a low-rank matrix completion framework is established to estimate the distances among all the unknown target nodes. After obtaining all distance measurements, a simple but robust approach based on Semidefinite Programming (SDP) relaxation is applied to estimate the location of target nodes. Compared with some existing algorithms, the proposed approach has three-fold merits: firstly, it locates multiple targets using less anchor nodes simultaneously; secondly, it shows more robustness to path model error and noise level; thirdly, it has higher localization accuracy than some existing methods. Simulation results illustrate the superiority of our proposed algorithm.