Experimental and performance analysis of non-line-of-sight (NLOS) localization methodology

• Main Author: Chan, Shi Mei
• Other Authors: Tan Soon Yim
• Format: Final Year Project
• Language: English
• Published: 2016
• Subjects: DRNTU::Engineering
• Online Access: http://hdl.handle.net/10356/69262
• Institution: Nanyang Technological University

As the Global Positioning System (GPS) does not work in the indoor environment due to strong signal attenuation and dense multipath effects. Thus, indoor localization system is necessary. Precise positioning in indoor environments faces challenges due to extensive multipath and non-line-of-sight (NLOS) conditions such as reflection from walls and obstruction of furniture between mobile device (MD) and reference device (RD). To solve the multipath and NLOS issue, traditional methods try to detect the NLOS nodes then discard them during localization, and using line-of-sight (LOS) nodes only to localize. However, this kind of method is unstable when the number of LOS RD is insufficient. Recently, some algorithms have been proposed by utilizing the multipath to assist localization instead of discarding them. This project focuses on analysing the performance of NLOS localization algorithm by utilizing the multipath to assist localization. The concept of virtual RD is introduced to consider the NLOS paths as virtual LOS paths which originated from the virtual RDs. Experiment has been conducted to collect the channel response data between transceivers, which are used to estimate the time of arrival (TOA). The channel is measured in the frequency domain using a vector network analyser (VNA). The frequency sweep is from 2GHz to 3GHz with 1601 frequency points. Secondly, the Space Alternating Generalized Expectation Maximization (SAGE) algorithm is used to estimate the TOA parameter using experimental data. Thirdly, using TOA-based NLOS localization algorithm to localize MD. Lastly, analysing the performance of the NLOS localization algorithm. Based on the error analysis, it can be shown that The TOA-based NLOS algorithm is able to reduce errors resulting from multipath conditions as it takes the NLOS path into consideration during localization.