Abstract: A low complexity joint channel estimative and decoding algorithm for STBC-OFDM systems in time varying or unknown channel is presented.The starting point of derivation of the algorithm is the observation that the least square cost function for channel estimate is equivalent to the one for standard STBC decoding,so the iterative method of cyclic minimization can be exploited for resolving them.The features of both the OFDM signal and the orthogonal structure of STBC can be exploited to reduce the complexity of resolving the two functions greatly.Duration the iteration,the signal is treated as a continuous variable other than a discrete variable,and then is adjusted using the finite alphabet property,which reduces the complexity and speeds up the convergence.Simulation validates the effectiveness of the algorithm.