Fuzzy Centroid Localization Scheme for Unbalanced Deployments of Wireless Sensor Networks

This paper proposes a novel methodology to mitigate the effect of unbalanced known nodes’ positions for location approximation in wireless sensor networks. In a practical deployment, some nodes may not properly be in uniform places, and perhaps, due to unequal power consumption of large-scale networks while performing sensing, computing, and transmitting tasks. Kmeans clustering is applied to select a representative of the known nodes where their positions are close together, and each of which will be then fed into fuzzy logic systems to determine a proper weight to finally use in the actual location determination process with weighted Centroid. The effectiveness of our methodology is evaluated via a large scale simulation with regard to node density, coverage, and topology, against a traditional Centroid, its fuzzy systems, and DV-Hop.