The application discloses a water acoustic
MIMO communication multi-channel different
carrier frequency offset two-stage joint
estimation method and relates to the technical field of water acoustic communication. In view of the problem that the
estimation performance is reduced due to model mismatch in the multi-channel different CFO scene, the method adopts a coarse-precise two-stage
estimation framework: in the coarse estimation stage, the minimum cost function value is searched to determine and compensate the multi-channel consistent CFO component through candidate CFO compensation, tentative channel estimation and
equalization; in the fine estimation stage, the multi-channel different residual CFO and the channel are jointly estimated based on the sparse Bayesian learning technology, and a low-complexity closed-
form solution is obtained through Taylor
series expansion by using the small characteristic of the residual CFO. The application does not need to configure zero subcarriers, effectively solves the model mismatch problem while maintaining
high spectral efficiency, reduces the estimation difficulty, improves the CFO estimation precision, and is suitable for actual water acoustic
MIMO-OFDM communication systems.