Signal processing technology of multi-user OTFS communication system based on spectrum cognition

In high-speed mobile communication scenarios, time–frequency dual-selective fading induces severe inter-symbol interference in traditional orthogonal frequency division multiplexing systems, drastically degrading communication reliability. Orthogonal Time Frequency Space (OTFS) modulation effectively avoids channel dynamic fading through delay-Doppler domain transmission, but still faces the problem of cross-domain interference and resource allocation in multi-user spectrum sharing scenarios. To this end, a spectral cognition-based MU-SISO-OTFS system signal processing technology is suggested by integrating signal detection optimization and intelligent power allocation techniques. First, grounded in the sparse quasi-cyclic structure of the OTFS channel matrix, a zero-forcing, minimum mean square error linear detection algorithm is designed, and a nonlinear iterative detection algorithm grounded in Message Passing (MP) is suggested. Secondly, in the underlay spectrum sharing mode, for bidirectional interference between the Primary User (PU) and the Secondary User (SU), an equal power allocation strategy under zero-forcing mechanism and a binary iterative power allocation algorithm with Successive Interference Cancellation (SIC) technology are designed. The outcomes reveal that the minimum mean square error detection block error rate is as low as 3.8 × 10 –4 at a signal-to-noise ratio of 25 dB, which is nearly an order of magnitude lower than the zero-forcing algorithm. The MP algorithm converges its bit error rate to 1.8 × 10 –4 with a relaxation factor of 0.6 and 8 iterations, taking only 12.4 ms. The total rate approximation of the suggested binary algorithm SU reaches 99.45%, and the execution time is only 8.35 ms. With the introduction of SIC, the reachable rate of SU at 40 dBm power is 15.1 bps/Hz and the outage probability is 0. Moreover, at a mobile speed of 500 km/h, the proposed SIC-assisted bisection scheme maintains a per-grid resource allocation fairness index of 0.84 while achieving an SU achievable rate of 11.2 bps/Hz, demonstrating a favorable balance between the system throughput and the resource allocation fairness in ultra-high-speed scenarios. The study provides a high-performance, low-complexity OTFS signal processing solution for high-speed mobile communication systems, which has significant engineering application value.