| IWith the rapid development of the mobile Internet and the Internet of Things,there is an increasingly urgent need for indoor location based services.Thus,the technology of indoor localization has attracted significant attention.Due to the integration of various sensor modules such as inertial navigation,geomagnetic,wireless transceivers and receivers,speakers,and microphones,the commercial off the shelf smart devices(e.g.,smartphones)become an important source of location information.Moreover,with the enhanced performance of acoustic sensors,localization via the ultrasonic signal,which is inaudible to human but can be identified by these smart devices has become a hot research topic and has been playing an important role in indoor localization systemsHowever,due to the limited transmission distance of the ultrasonic signals the inability of smart phones to match the received signals with the beacons,most of the existing smartphone based ultrasonic location methods can only be applied in small-scale application scenarios.There is a lack of an effective solution for high-performance localization in large-scale scenarios,where the signal transmission distance usually needs to exceed 20 m and a massive number of beacons need to be deployed.The main contributions of this dissertation can be summarized as follows(1)In order to research the main challenges in the implementation of the indoor ultrasonic localization system in the large environment,we study the acoustic propagation mechanism,sort out the environment interferences,establish an indoor ultrasonic simulation platform and localization system simulation model for feasibility analysis and solution design(2)Considering the frequency,length,energy and waveform of the LFM signal and the frame length,coding number and orthogonal performance of the localization system,we design an encoded signal with anti-noise ability,high precision time estimation and strong orthogonal performance,to discriminate between beacons(3)To alleviate the environmental interferences(e.g.,noise,multiapth,NLOS,movement),an improved decoding scheme is proposed.We detect the weak signal by normalizing the amplitude,estimate the arrival time by searching the first path signal,merge the environment information into reference signal to decode the received signal and compensate the time shift caused by doppler effect.(4)We do extensive experiments in a long corridor,where the environment interferences are serious.The results demonstrate that the signal can be detected within 50 m range and guarantee 25 cm ranging accuracy in the multipath environment.In addition,the improved decoding scheme achieves more than 30 percent success rate than the based one in the serious NLOS condition and increases 10 percent success rate by adjusting the search range in the movement condition. |
