This disclosure relates to methods, systems, and devices for AI/ML assisted wireless channel fingerprinting, estimation, and prediction. In one example embodiment, a method of combined AI/ML assisted wireless channel fingerprinting and channel prediction is disclosed. The method includes using a trained neural network to fingerprint the channel with the channel fingerprinting results advantageously being leveraged to improve the channel prediction.

A near-optimal Karhunen-Loeve basis expansion modeling (KL-BEM) orthogonal time frequency space (OTFS) receiver with superimposed pilots has been proposed for high-mobility communications with Doppler spread channel. First, an initial KL-BEM channel estimation is conducted by superimposed pilots, followed by the removal of superimposed pilots from the received OTFS signal and equalisation by message passing (MP) algorithm. After that, the detected data symbols are utilized as pseudo pilots along with the superimposed pilots to refine both KL-BEM channel estimation and equalisation in an iterative manner. Simulation results confirm the superior performance of the proposed KL-BEM OTFS receiver over the prior art in terms of the mean-square-error (MSE) of channel estimation and bit error rate (BER). It also has a close BER performance to the BER lower bound obtained by assuming perfect channel estimation. It contributes to high spectral efficiency and fast convergence performance.