本文对线性DPCM(差分脉码调制)编码的误码问题进行了系统的分析,从误码响应出发,讨论了误码图形的稳定收敛问题,用误码能量作为评价误码响应的主观效果、误码能量的有关性质、误码能量与预测系数、取样点位置及点数的关系,以及误码能量最小化等问题,为预测方案的选择提供了一定的依据。 This paper systematically analyzes the bit error problem of linear DPCM (Differential Pulse Code Modulation) coding. From the bit error response, we discuss the problem of the stable convergence of bit error code, and use the bit error energy as the subjective effect to evaluate error response. The nature of the error energy, the relationship between the error energy and the prediction coefficient, the position of the sampling points and the number of points, and the minimization of bit error energy provide the basis for the selection of the prediction scheme.