Iterative decoding algorithms for Reed-Solomon (RS) product codes have been proposed. In one of the iterative decoding algorithms, a soft-input iterative bounded-distance decoding algorithm is applied to the constituent codes. The bounded-distance decoding corrects up to t{sub}0 (＝[(d{sub}(min) - 1)/2]) where d{sub}(min) is the minimum distance of a RS code, and generates at the most one candidate codeword. And then soft-output values are calculated using the candidate codewords generated in the iterative bounded-distance decoding algorithm. The soft-output values are used as soft-input values at the next decoding stage. Because the number of the generated candidate codewords is not large, the quality of the soft-output values is not good enough. This fact causes that the iterative decoding algorithm can not achieve good error performance. In this paper, we study a soft-input iterative bounded-distance+1 decoding algorithm for RS codes in which bounded-distance+1 decoding that corrects up t{sub}0 + 1 errors and generates candidate codewords with a relatively large number, is performed. Simulations were made to evaluate the error performances, the average numbers of generated candidate codewords and the quality of soft-output values over additive white Gaussian noise channel using binary phase shift keying modulation. From the results, we show that the iterative bounded-distance+1 decoding algorithm is better than the iterative bounded-distance decoding algorithm.
展开▼
