Theoretical Gaussian Mixture Modelling for Cooperative Localization in Wireless Sensor Networks

Abstract: This work provides one theoretical basis for cooperative localization in Wireless Sensor Networks (WSNs) with Gaussian Mixture Modelling (GMM). The work attempts to solve the problem of sensor node positions determination in unknown and hostile environment by using probabilistic clustering methods. By analysing the synthetic data and simulating studies, the proposed GMM based approach is shown to outperform traditional range-based methods in terms of the localization accuracy, with lower Root Mean Square Error and the ability to mitigate the noise, signal interference and uncertainties of the wireless channels.

Keywords: Cooperative Localization, Wireless Sensor Networks (WSNs), Gaussian Mixture Modelling (GMM), Sensor Node Positioning, Probabilistic Clustering, Localization Accuracy, Root Mean Square Error (RMSE)