As Internet of Vehicle (IoV) systems evolve, the demand for real-time high mobility user positioning is escalating. High-speed vehicle positioning using GPS faces significant challenges in urban and canyon environments due to signal blocks and high deployment costs. While ground-based wireless networks offer a promising alternative, conventional delay- or angle-based positioning methods often require wide bandwidths or complex antenna arrays to achieve sufficient resolution. Recently, Orthogonal Time-Frequency Space (OTFS) modulation has emerged as a promising solution for next-generation wireless communication, offering robustness against Doppler effects and enabling full channel diversity. By operating in the delay-Doppler (DD) domain, OTFS provides a compact and sparse representation of rapidly varying wireless channels, making OTFS particularly well-suited for positioning and sensing applications.

Particularly, fractional doppler has traditionally been regarded as a challenge in OTFS systems due to the additional Doppler spread it introduces. In the first phase of our research, we approached fractional Doppler from a new perspective, recognizing its potential for high-resolution Doppler-only positioning and adopt it in high-speed railway scenarios for positioning along with velocity detection. Specifically, correlation-based Doppler estimation—commonly used in OTFS channel estimation—is employed to obtain fine-resolution fractional Doppler measurements. The Doppler-only positioning problem is then formulated as a nonlinear equation and solved using the Halley method. Simulation results demonstrate that the proposed approach, even with standard bandwidth similar to commercial communication systems, outperforms both integer-Doppler-based positioning and ultra-wideband (UWB) time-of-arrival (TOA) methods. Moreover, as only Doppler shifts are involved instead of delay or any other angle-related measurement, the positioning scheme can effectively work without massive antenna arrays or large bandwidth.

The second stage of this project aimed to enhance the OTFS-based vehicular positioning by integrating intelligent refection surfaces (IRSs), a cost-effective passive antenna technique acknowledged for improving multipath interference and mitigating non-line-of-sight channel effects. Combining continuous phase modulation with beamforming allows simultaneously control of the Doppler shift and departure angle of the reflected signals. This approach assigns different Doppler shifts to each reflected signal, facilitating source identification within the OTFS channel. For precise positioning, this research converted the fractional delay-Doppler estimation challenge into a super-resolution two-dimensional frequency extracting problem in the time-frequency domain. Coupled with the 2D spectrum estimation method, accurate user delay and Doppler information can be further extracted. Numerical evaluations demonstrated that employing IRS as a spatial anchor can facilitate estimating users' multipath positions through a single signal transmission, significantly enhancing overall convenience.