An SDR-Based Evaluation of Maximum Likelihood Estimator and MUSIC Algorithm for TDOA in 5G Campus Networks

In today's research on Fifth and Sixth Generation Mobile Networks (5G and 6G), a strong focus is placed on environmental sensing, including the localization of User Equipment (UEs). This study explores Time Difference Of Arrival (TDOA) estimation based on the Maximum Likelihood (ML) estimator and the Multiple Signal Classification (MUSIC) algorithm. A test bed based on an Software Defined Radio (SDR) platform was developed for evaluation purposes, and Uplink (UL) TDOA measurements using 5G Sounding Reference Signals (SRSs) were conducted in an indoor environment within the frequency range of 5G campus networks. Various factors, including the number of antennas, transmitter-receiver distance and Signal to Noise Ratio (SNR), were analyzed. In some measurements, the ML estimator exhibits higher error than the MUSIC algorithm, primarily due to multipath components and the ML estimator's inability to resolve them temporally. Comparing the pseudospectra, the MUSIC algorithm demonstrates superior temporal resolution compared to the ML estimator. However, this advantage is not observed across all measurement series. In low SNR conditions and with an increasing number of antenna elements, the ML estimator outperforms the MUSIC algorithm.