Greedy Probabilistic Approach for Localization in IoT Context

In this paper we propose a greedy probabilistic approach for localization in Wireless Sensors Network (WSN). This topic has received much attention since the WSN are considered as the basis in the emerging area of Internet of Things (IoT). The proposed method aims at increasing the performance of the grid based Compressed Sensing (CS) localization algorithm. This latter is based on the sparse nature of localization problem and select one grid point as user position. The grid point is selected based on correlation property. We propose in this paper to select a grid point based on probabilistic approach where grid point probabilities are calculated from the received signal strength. In a second step we propose to combine the grid positions weighted with their probabilities. The performance of the proposed approaches is evaluated through simulations and compared to CS algorithm results.