Tracking of Moving Objects in a Wireless Sensor Network with use of Kalman Filtering and Machine Learning

WSN encompasses a significant number of sensor nodes that were generally fortified by sensing, processing as well as communication competences. Such equipment’s, gather conservational data using their sensors, besides, send such information to the base station. Considering that in these types of networks there is restricted electricity capacity and it is usually impossible to revitalize or substitute dead nodes; therefore, energy management is of considerable significance at such types of systems. Different methods are suggested and proved for tracking the targets in such a network using various techniques as well as evolutionary algorithms. Many types of research have tried to improve the performance of each of these methods. Among the presented methods, methods that can estimate the subsequent position of the goal at the next period and in terms of specific techniques are more efficient. By predicting a target move, only a portion of the network nodes should be involved for tracking in the next period, resulting in a significant reduction in energy consumption. In this thesis, a new method is presented for target tracking with the use of Kalman filtering and machine learning. The proposed method results in a network with better performance and less energy consumption. Keywords: Kalman filtering, WSN, Estimation and prediction, Target Tracking.