Search

In this paper, we consider an interieave-division multiple-access (IDMA) system with multiple users transmitting over a shared additive white Gaussian noise channel. In an IDMA system, the information sequence of each user is encoded by a low-rate serially concatenated code consisting of a convolutional code and a repetition code. To reduce decoding complexity, it has been proposed to only decode the inner repetition code for the first few iterations before including the full concatenated code constraints into the decoding process. Here, we characterize this concept of switched decoding as the parallel decoding of two complementary punctured codes and propose a systematic design procedure based on the extrinsic information transfer chart and trajectory analyses, leading to the best decoding convergence-complexity tradeoffs. Index Terms--Iterative decoding, multiple access, multiuser decoding, transfer function analysis.

A study proposed an effective approach for reducing blocking effects in transform image coding. The method requires two projection operators and the theory of projection onto convex sets (POCS) to ensure convergence of the iteration. However, the algorithm that was used is not strongly related to the theory of POCS. An alternative formal basis for establishing convergence and a more flexible theory for explaining the potential applications of the algorithm are presented.