Traditional Time of Flight (TOF) estimation methods [1] are inaccurate in multipath and noisy environments. This paper enhances TOF measurement with the Multiple Signal Classification (MUSIC) algorithm [2]. It uses spatial smoothing on channel state information (CSI) dataset, frequency domain analysis and MUSIC-based methods to distinguish signals, and the least squares method to reduce multipath-induced errors for transmitter positioning. It also conducts linear fitting, maximum likelihood estimation, and statistics to evaluate and reduce TOF errors. The paper achieves precise positioning in multipath scenarios and improves the MUSIC algorithm’s accuracy and applicability under noisy conditions by analyzing error sources.