Source nodes localisation algorithm for large-scale wireless sensor networks using self-organising isometric mapping

An efficient, accurate and distributed algorithm for nodes localisation in large scale sensor networks (WSN) is presented, which uses the signal strength values received by the sensors as measurements and determines the locations of each sensor. At first, the algorithm divides a large scale WSN into a number of small clusters according to a clustering algorithm; then, a local map is constructed in each cluster by using the SIEMAP algorithm locally. Finally, all local maps are merged together to get a global map according to a merging algorithm. The global map can be optionally further refined for higher accuracy. Simulations demonstrate that the proposed method is robust to measurement errors and can achieve comparable results when fewer anchor nodes are available than previous methods.