Improving mobile node tracking performance in NLOS environments using cooperation

It is well-known that non-line-of-sight (NLOS) propagation degrades the accuracy of localization in indoor and harsh environments. In such environments, source nodes either do not have connections with a sufficient number of anchor nodes due to strong attenuation or have very poor range estimates due to NLOS propagation. Cooperative localization, in which a source node communicates with both anchor nodes and other source nodes, can benefit such scenarios by providing additional measurements that can significantly improve the localization performance. While most of the earlier works have studied static node localization in noncooperative networks, cooperative tracking of mobile nodes in NLOS environments is still an open problem. To address this open problem, in this work we develop an technique for cooperative tracking applicable to NLOS environments. More specifically, we develop a novel node tracking algorithm based on semidefinite programming (SDP) which has the ability to mitigate NLOS propagation. Our simulations show that the proposed SDP estimator outperforms the classic extended Kalman filter for cooperative networks in NLOS environments.