Three-Dimensional Localization in Anisotropic Wireless Sensor Networks Using Fuzzy Logic System

Precise node localization is one of the critical issues for many wireless sensor network (WSN) applications. Existing localization schemes in WSN mainly focus on two-dimensional (2D) plane, while in real practice; nodes are usually deployed in three-dimensional (3D) space. There are still many issues in 3D localization such as high computational complexity, low localization accuracy, low positioning coverage, and relying mainly on anchor nodes. On addressing the issues of current 3D localization algorithms, we propose two range-free localization algorithms for 3D space in anisotropic environment using application of fuzzy logic system (FLS). To evaluate the performance of proposed algorithms in practical scenario, anisotropic property of nodes is considered. In proposed methods, only received signal strength (RSS) information between target nodes and their neighbouring anchor nodes is sufficient for estimating target nodes locations. The RSS information gives clue to find out the distances between them. To overcome the nonlinearity between RSS and distance, edge weights between target nodes and their neighbouring anchor nodes are considered to estimate the positions of target nodes. Further to model these edge weights; we use FLS in this paper. TLBO (teaching learning based optimization) and M-IWO (modified invasive weed optimization) techniques are used to further optimize the edge weights separately to achieve the better accuracy. The simulation results of the proposed schemes are better as compared to centroid method, weighted centroid and some existing 3D localization algorithms in terms of location accuracy, stability, positioning coverage and scalability.