OTDOA-Based Positioning in Narrowband IoT

Abstract: This thesis considers the problem of improving the performance of observed time difference of arrival (OTDOA) based positioning for Narrowband Internet of Things (NB-IoT). A new iterative low-complexity algorithm based on expectation maximization successive interference cancellation (EM-SIC) is presented. The algorithm tackles the near-far problem by using an iterative SIC, where for each received signal a frequency offset compensation and channel estimation is performed. A low receiver sampling rate of 1.92 MHz is assumed and the time of arrival (TOA) estimates are refined by interpolation of correlation results to a higher sampling rate. At the final stage of the algorithm, the problem of multipath fading is addressed using an iterative first-path detector. It is shown that the proposes EM-SIC algorithm greatly increases positioning performance compared to conventional methods and the estimated TOA variance almost attains the Cramér-Rao lower bound.