Simplified OTFS-based IDZT-DZT-LMMSEE for robust high-doppler channel estimation in V2X-joint RADCOM system

This paper addresses the high mobility channel issue for vehicle-to-everything (V2X) joint Radar-Communication (Radcom) systems, a phenomenon that is commonly referred to as high-Doppler channel estimation. In high-Doppler channels, coherence time is reduced by rapid channel variations. V2X Radcom systems are subject to multiple reflections of the transmitted signals, with vehicle motion causing high-Doppler multipath shifts. This results in residual Inter-Symbol Interference (ISI) and, as a consequence, Inter-Carrier Interference (ICI). In this paper, we introduce a Simplified Orthogonal Time-Frequency Space (S-OTFS) modulation based on an Inverse Discrete Zak Transform (IDZT) and Discrete Zak Transform (DZT), joined with a Linear Minimum Mean Square Error Equalizer (LMMSEE), denoted as IDZT-DZT-LMMSEE. This approach is proposed as a solution to the challenges posed by V2X communications in a Radcom system, which utilizes the Orthogonal Frequency Division Multiplexing (OFDM) and standard Orthogonal Time-Frequency Space (OTFS) modulations. These modulations are susceptible to encountering challenges in high Doppler channels due to their inability to effectively cancel the ICI phenomenon and processing latency. The S-OTFS joint Radcom based on IDZT-DZT-LMMSEE approach exhibits superior ICI suppression capabilities up to ~ 85% in the delay-Doppler (DD) domain in high mobility channels with a relative Doppler shift up to 2162 Hz in the presence of multi-path scenarios. The innovative Radcom V2X system demonstrates superior performance, as evidenced by the experimental results. The framework was validated based on tests involving three vehicles operating at different high speeds. Compared to standard OTFS techniques and OFDM variants, the proposed approach achieves a highly desirable 11.40% reduction in error vector magnitude root mean square (EVM RMS) and a 2 × 10⁻⁴bit error rate (BER) reduction with 32-QAM modulation at a fixed SNR of 25 dB, and a reduced peak-to-average power ratio (PAPR) of 7.8 dB. Real-world speed scenarios using the MATLAB software development framework provide convincing evidence that the proposed S-OTFS approach is an effective solution for high-speed vehicle detection and communication, even in the presence of high Doppler channels.