Interference approximate method (IAM) for orthogonal frequency division multiplexing / offset quadrature amplitude modulation (OFDM/OQAM) channel estimation needs the value of preamble at the background of un-cooperative communications. Motivated by this consideration, an OQAM characteristic based-IAM (OCB-IAM) algorithm, which used IAM preamble framework and OQAM symbols’ real finite-alphabet characteristics, was introduced. OCB-IAM algorithm estimated the amplitude and phase of channel fading coefficients separately, and just needed preamble position, which was a kind of semi-blind channel estimators. OCB-IAM algorithm converted white Gaussian noise into single tone interference based on the secondorder statistics of received symbols. Then it could be proved that OCB-IAM algorithm has a better performance than IAM algorithm in medium and low SNR regime. Numerical results show the validity of theoretical analysis and the reliability of OCB-IAM algorithm.