Research On Wireless Sensor Network Localization Based On Ir-UWB Technology

The application prospect of wireless sensor network is extremely widespread, and thenode localization is its application premise. IR-UWB is one kind of emerging wirelesscommunication technology, which uses a nanosecond-narrow pulse to transmit theinformation and has fine time resolution, and is able to achieve some gigahertz in bandwidthwithout carrier modulation. IR-UWB has a perfect ranging accuracy and low hardwarecomplexity, which lead itself to extremely suit in wireless sensor network physical layertechnology. In IR-UWB wireless sensor networks, to fully use the strong time resolution ofIR-UWB, TOA based localization algorithm is often used.The thesis systematically investigates key technology of IR-UWB ranging module inwireless sensor networks. First the research background and the significance of the issue areintroduced, and the IR-UWB wireless sensor network technology is analyzed. Second, thesignal of IR-UWB system, pulse modulation and channel model, standard work are presented.The existing wireless positioning technology and the localization algorithm based on the timeof arrival are studied in depth. Finally, the TOA estimation algorithm is studied and itsperformance is simulated and compared. In view of the existing TOA estimation algorithm,three kinds of new TOA estimation algorithm are put forward, and the main innovations areas follows:1. According to the characteristics of the MMR-TC algorithm that theperformance of long time integral cycle is worse than short integral time period TOAestimation, a novel two step TOA estimation algorithm based on energy detection isproposed.Simulation result show that the ranging accuracy has a obvious promotion, and itsperformance can basically achieve the short period energy detection algorithm, but thealgorithm complexity, time complexity and space complexity are much smaller than the shortintegral period energy detection algorithm.2. A new NMMR-TC algorithm which used theratio of maximum energy block to mean energy block to set the normalized threshold isproposed. The algorithm is simulated and the result shows that the performance of thisalgorithm is almost the same with MMR-TC algorithm.3. For NMMR-TC algorithm is not better than MMR-TC at the high signal to noise ratio, combined with the new proposedMEAN-TC algorithm to be merged into the MAX-MEAN-TC algorithm. Simulation resultsshow that the algorithm achieves the better performance.