In this paper, we are concerned with developing a common implementation architecture for both turbo codes and LDPC codes. We view the constituent convolutional code in a turbo code as a block code, and construct a sparse parity check matrix for i