Research On Wireless Localization In Indoor Environment

In recent years, context-aware service is being turned from dream into reality due to the development of the computer technology and the wireless communication technology. The position information of customers is one of the most important environment parameters. The traditional GPS(Global Positioning System) and cellular networks systems have the abilities to provide accurate localization and navigation services for outdoor users. When it comes to indoor conditions, the wireless signals get attenuated, slows down and get dispersed because of the influence of furniture, persons, walls and other obstacles, so the traditional outdoor localization techniques will cause large errors when used in indoor environment. Therefore, the indoor localization techniques become current popular research.There are mainly two reasons that affect the precision of indoor localization. The one is NLOS(Non Line of Sight) propagation, the other is multipath propagation. In order to solve the above two problems, this dissertation studied from three aspects: wireless channel modeling, location parameters sensing and localization algorithms. It also proposed the corresponding solutions and demonstrated their effectiveness though algorithms comparisons and simulations. The major innovations of this dissertation are summarized as follows:1. The information of NLOS and multipath which affect the localization precision can be utilized. The geometric propagation model for indoor wireless signals based on TOA(Time of Arrival) and DOA(Direction of Arrival) measurements was proposed according to the known floor plan by establishing virtual stations which contains the situations of direct propagation, transmission, reflection and diffraction. By doing this, the NLOS problem was turned into LOS(Line of Sight) problem. It can deal with both one bound scattering and multi bound scattering, so it can apply for more severe NLOS conditions. Besides, a two steps weighted least square algorithm based on TOA and DOA measurements of virtual paths was proposed to estimate the location of the target. Simulation results showed that proposed method achieved high accuracy in different NLOS conditions, especially the ablility of dealing with multiple scattering paths was better than the existing.2. Antenna array is necessary for the DOA estimation, so the system hardware structure is complex and computational complexity is high. In the situation that the signal experiences one reflection or diffraction at most between the base station and the mobile station, the TOA indoor wireless signal transmission model based on the virtual station was proposed. Since this model only needs TOA measurements, both the system hardware structure and the computational complexity are simpler compared to the existing related geometric model. The simulation results showed that the proposed method achieved high positioning accuracy in the NLOS environment with one reflection or diffraction.3. The performances of the above two channel models were analyzed. The CRLB(Cramer-Rao Lower Bound) and GDOP(Geometric Dilution of Precision) of the two models were deduced.4. The FrFT-IAA algorithm was proposed for the DOA estimation of wideband chirp signals with small amount of snapshots. The proposed method extended the IAA approach which was apllied to the DOA estimation for narrow band signals to that for wide band signals using FrFT(Fractional Fourier Transform). The energy-centrobaric spectrum correction method was also introduced to correct the results of the DOA estimation. The simulation results showed that both the accuracy and the resolution of the DOA estimation of the proposed method are high with only a few snapshots.