M.Sc. Tezi Görüntüleme

       Wireless sensor networks are created for using low cost, low power sensor nodes to collect useful information from a targeted environment in various applications. Location estimations are one of the areas where these networks can be exploited. The main aim of this study is to control a fleet via wireless sensor networks that only leader vehicle has controlled by a human being. Instead, some of the vehicles can be equipped with GPS receivers and the rest of fleet can find their own position via RSSI (Received Signal Strength Indication), TOA (Time of Arrival) and TDOA (Time Difference of Arrival) methods. Based on our MATLAB based simulations, we analyse the relationship among various factors such as GPS (beacon) node number and distance etc. Simulation results show that location estimation errors can be significantly reduced with right beacon node number selection and correct positioning of nodes.

       There are three GPS integrated vehicles in this experimental study. Naturally, location informations of these vehicles will be known. Moreover another vehicle can estimate its location according to those three vehicles via Trilateration method with the help of RSSI technique. Cramer Rao, Tulip and Least Square methods will be applied to trilateration and results of those methods will be compared. Communication quality, calculation time and speed synchronization will be other factors which affects in experimental study in addition to simulation factors.

       Key Words: Wireless Sensor Networks, RSSI, TOA, TDOA, Localization, Cramer Rao, Tulip, Least Mean Squares.