Research And Implemention Of Localization Algorithms In Mixed LOS/NLOS Environments

With the development of wireless techniques,wireless network localization becomes one of the most important alternating solutions to the position information,when the global positioning system(GPS)fails in indoor or dense canyon scenarios.Due to the fact that the occurrence of mixed line of sight(LOS)and non line of sight(NLOS)of the indoor environments,we present an interactive multiple model(IMM)based localization and navigation framework by which the NLOS ranging errors can be obviously mitigated.In detail,the main contributions of this thesis can be concluded as follows:First of all,our thesis studied how to eliminate errors under the background in mixed LOS/NLOS environment when Target Node(TN)is moving.We established position system by wireless sensor networks(WSN)which distance measurement based on Received Signal Strength Indicator(RSSI).A traditional NLOS error eliminate method usually need to identify NLOS firstly.But the NLOS identification accuracy has limited this method improved positioning accuracy.Our thesis can avoid the shortage of the traditional method by use IMM model.Then,we extended IMM's model to overcome a variety of obstacles environment.Due to different obstacles cause different signal attenuation in the indoor environment,extended IMM algorithm can adapt to the complex indoor environment effectively.Finally,we have obtained the NLOS error model by measuring and statistic different obstacle in our experiment place.After obtaining these measured value based on ZigBee module,we use it in extended IMM algorithm.We have further studied the impact of the model number of extended IMM algorithm on localization.Generally,we tried to present novel and effective strategies to eliminate ranging and localization errors introduced by various obstacles.A multiple state IMM model is adopted to solve the NLOS problem.Both theoretical and practical results are illustrated,to show the performance improvements achieved by the presented strategies.The presented frameworks,algorithms and numeric results are of valuable to the navigation in typical LOS/NLOS mixed environments,and applications in similar scenarios.