TDOA Localization Algorithms In Wireless Sensor Networks

With the advances in the modern microelectronic technology, Micro-Electro-Mechanism System (MEMS), system on a chip (SoC), nano-materials, wireless communications, signal processing, computer network technology, etc, as well as the rapid development of Internet, the technology of obtaining the wireless sensor network node location information develops from an independent single-mode to integration, miniaturization and then to the direction of intellectualization, internetization, thus it is becoming one of the most important and most basic technologies in wireless sensor networks (WSN). This paper will first introduces the development prospect of the wireless sensor networks and the overseas and domestic research status of localization algorithm, then summarizes the existing typical positioning technology and localization algorithm in the literature. TDOA ranging technology has simple principle, needs no additional hardware facilities, as well as the perfect ranging effects in the LOS environments, so it is widely used in the positioning of wireless sensor nodes. This paper compares the classic Chan TDOA algorithm and Kalman filter-based TDOA reconstruction algorithm, then finds the Chan-style algorithm can achieve relatively high positioning accuracy only in the environment of the Gaussian noise, but in the non-Gaussian environment, the algorithm’s positioning accuracy is greatly reduced. The Kalman filter-based TDOA reconstruction algorithm gets a certain improvement compared with the Chan algorithm, but in the prediction and recursive process it considers that changes in the location information of the wireless sensor node is a process of mutation, while in real wireless sensor networks scenarios, the location information of nodes is a gradual process. Aimed at overcoming the shortcoming of the Kalman filter-based TDOA reconstruction algorithm, this paper proposes an improved Kalman filter algorithm, by adding excessive state during the time interval between the two changes in the position of the node in order to overcome the error due to the mutation. Finally, this paper uses the MATLAB simulation tool to simulate the positioning results of the three algorithms, also evaluates the CRLB indicators of the three algorithms. Simulation results show that positioning performance of our improved algorithm is higher than the other two algorithms.