Researches On Ranging And Localization Algorithms For Ultra Wide Band Sensor Networks In Special Environments

Localization, search and rescue operations in mines and other special environments have recently become an urgent issue after series of disastrous events of underground mining. The application of wireless sensor networks provides a dynamic and fast method of networking in support of underground communications, and Ultra Wide Band (UWB) have emerged as the promising technology for mining environment due to its abilities on low power consumption, high data rate, anti multipath and interference, accurate ranging and localization, etc. Therefore, UWB sensor networks provide a feasible solution for underground communications and localization. The thesis mainly researches on ranging and localization algorithms of UWB sensor networks in special environments.First of all, the thesis introduces the concepts of UWB, its advantages and the state of the art. Then, we analyze and compare the ways of generating UWB pulse and signal, several types of channel model, and proposed algorithms of ranging, localization and meida access. Feasible schemes for underground UWB communcations and localization have been advised after above discussion.We choose Time Hopping UWB (TH-UWB) for underground communication because of its high performace-price ratio and TH code that can resolve multiuser interference. According to the characteristics of underground TH-UWB signal propogation, we refer to channel models of IEEE 802.15.4a and modify the corresponding indoor modeling. An indoor alike underground UWB path loss model is proposed, which takes into accout the different shading on varied areas and locations. On the basis of the propagation model, maximum likelihood estimation based Received Signal Strength Indication (RSSI) is designed to improve the stability of underground RSSI ranging. The Cramer-Rao lower bound is derived and then used to verify the algorithm of RSSI raning in simulations.Underground localization of UWB sensor networks is based on the underground channel model and RSSI ranging. Through analysis of the relation between localization error and ranging error, the thesis merges the ranging into localization process and provides an underground localization algorithm by modification of the least mean square error method. The localization algorithm mitigates the dependence on the number of reference nodes, and gurantees the accuracy of localization. Futhermore, we utilize the advantage of multiple acces in TH-UWB, and take a method, named as Time Hopping-Code Division Multiple Access (TH-CDMA), for media access and multiuser communication. The method decreases the interference of channel access by multiuser, increases the throughput of ranging information, and improves the localization algorithm. Simulation results indicate that the underground localization algorithm can achieve a significant improvement on the accuracy of localization, and satisify the location-oriented service in mine and other special environments.