A Range-Free Node Localization Method for Anisotropic Wireless Sensor Networks With Sparse Anchors

In sensor networks with anisotropic properties, including irregular layout sensors, poor network connectivity and so on, employing traditional localization methods may lead to poor localization accuracy. Constructing a novel localization model suitable for anisotropic properties may effectively address the above problems. Generally speaking, anisotropic properties can lead to an increase in the hop count between anchor nodes and unknown nodes. While according to range-free localization principles, a higher hop count between anchor and unknown nodes typically results in poorer localization accuracy of unknown nodes. Inspired by this observation, we propose a weighted range-free localization model where the confidence level of distance between an anchor node and an unknown node is set to be inversely proportional to the hop count. Furthermore, within the proposed weighted model, we also introduce a rational distance estimation related to types of anchor pairs. Despite the model’s inherent nonconvexity and nondifferentiability, it can be reformulated into an optimization model with convex objective functions and nonconvex constraints through matrix transformations. To resolve these nonconvex constraints, we employ a successive convex approximation (SCA) algorithm that tightens nonconvex constraint to its first Taylor expansion version, leading to an iterative solver of original optimization problem. Simulation results validate that our proposed algorithm substantially improves stability and accuracy in estimating range-free node locations.