Incremental LLE for Localization in Sensor Networks

Received signal strength indicator (RSSI) gives a coarse initial measure of the inter-node distance at a low cost without the need for additional equipment or complexity. This necessitates the need for a mechanism to obtain accurate node locations from the noisy RSSI distance estimates. In this paper, an iterative nonlinear manifold learning technique, incremental locally linear embedding (ILLE), has been proposed for accurate node localization. The ILLE considers the one-hop neighborhood around the anchor nodes, as a reference structure. This structure grows iteratively to localize all the remaining sensor nodes in the network. Simultaneous localization mechanism further reduces the computational complexity of localization. Experimental results show that the ILLE is able to localize the nodes accurately in both normal and simultaneous scenarios. The ILLE is found to have higher accuracy in the typical scenario as compared with the simultaneous scenario. Results also indicate that the ILLE is able to localize sensor nodes with an increased accuracy of around 12.36% as compared with the centralized LLE and also outperformed other existing similar localization techniques.