In recent work on decoding space-time codes

it is assumed that a channel estimate is obtained.To estimate the channel increases the computation load and a loss in performance of decoding is incurred by the error of channel estimate.In this paper

we propose a new decoding algorithm for space-time block coding.Combined with the structure of space-time block coding

the subspace of the transmitted signal are exploited to carry out the decoding without knowledge of the channel.Simulation results show that with the increase of the precision of noise subspace

the performance of our decoding algorithm approaches that of the coherent decoding algorithm with knowledge of channel.