Efficient RSS-based collaborative localisation in wireless sensor networks

This paper presents a new collaborative location estimation method for wireless sensor networks WSN, referred to as an iterative tree search algorithm I-TSA. The proposed method is based on the grid search least square estimator LSE, which provides efficient estimation in the presence of noisy received signal strength RSS range measurements. The complexity analysis of the I-TSA algorithm showed that the computational requirement by each unknown-location sensor node scales linearly with the number of its neighbouring nodes, and that only a small communication overhead is required until its location estimate converges. This, in contrast to centralised methods, such as maximum likelihood estimator MLE and multidimensional scaling MDS, provides a feasible solution for distributed computation in large scale WSN. Furthermore, the performance of I-TSA, is evaluated with reference to the Cramer-Rao bound CRB and compared with MLE, MDS and MDS-MLE methods. The results showed that I-TSA achieves lower standard deviations and biases for various simulation scenarios.