AbstractThe performance of underlying Hamming codes in M‐QAM digital radio systems is analyzed in the presence of nonlinear distortions.An effective soft‐decision decoding algorithm is constructed to reduce the number of errors induced by the nonlinear and noisy channel.Then the residual bit error probability is calculated from the statistical estimations of the detector input samples. It is shown that for weakly and moderately nonlinear channels error correction coding can significantly improve the system performa