Abstract: Blind identification of channel coding is significant for both AMC in cooperative communications and passive listening in a non-cooperative context. Whats more, blind estimation and detection of interleaver parameters is a key step to recover original data sequence. In this paper, we study blind identification of three interleavers block interleaver, convolutional interleaver and helical interleaver, and make some comparison of them. Combining rank criteria with Gauss-Jordan Elimination through pivoting (GJETP) algorithm, blind identification of interleaver parameters in high bit error rate is realized and we discuss the depth, width, synchronization, and step of interleaver respectively. Based on deinterleaver theory and some soft information, a universal method for blind identification of interleaver is raised, which enhances the algorithm ability for adapting to real channel. The experiment indicates that the algorithms are effective and feasible.