MAP decoding of MTR codes in LDPC-MTR encoded magnetic recording systems

A MAP algorithm for decoding maximum transition run (MTR) codes, using the soft-decision approach, is presented in this paper. The algorithm is based on the Bayesian algorithm, and computes a posteriori log-likelihood ratio (LLR) for output bits, using codewords of the MTR code. The log MAP and maxlog MAP version of the algorithm have been developed, showing performance comparable to that of the original version. MAP algorithm offers 2dB coding gain, in combination with LDPC, comparing to the uncoded system, while utilization of the MTR, the 25% decreasing in overall required operations of detector over E²PR4 magnetic recording system.