Modified decoding algorithm for IRA code in MIMO-OFDM communication system

Irregular Repeat Accumulate(IRA) code has drawn a lot of attention recently to its prominent decoding algorithm, whose performance approaches to Shannon limit. Belief propagation decoding algorithm is the best decoding algorithm for IRA code, which has great hardware complexity when implemented. Min-sum decoding algorithm modifies belief propagation decoding algorithm at the cost of decoding performance. A modified decoding algorithm is proposed to simplify belief propagation decoding algorithm with a little of loss in decoding performance according to minimum mean square error rule. The modified decoding algorithm can be applied to MIMO-OFDM communication system to decrease Inter-Carrier-Interference. Simulation results show that if MIMO-OFDM system needs SNR to be 10⁻⁵, the encoding gain of modified decoding algorithm can be 0.25dB and 0.15dB better than normalized decoding algorithm and offset decoding algorithm respectively. And the modified decoding algorithm also has lower hardware complexity.