模拟信号转换成数字形式,必须要经过量化才能实现。标量量化的方法包括均匀量化和非均匀量化两种,这两种方法都会带来精度的损失。文中探讨了模拟信源量化的最优算法,并用C语言实现了高斯信源在不同量化电平下均匀量化和非均匀量化的最优步长的数值计算。通过对计算结果进行比较,给出了非均匀量化优于均匀量化的具体结论。-Analog signals into digital form, you must go through quantification can be achieved. Scalar uniform quantization methods include two kinds of quantitative and non-uniform quantization, these two methods will bring about the loss of accuracy. This paper discusses the analog source to quantify the optimal algorithm, and use C language implementation of the Gaussian source at different quantization levels under the uniform quantizer and the optimal non-uniform quantization step size of the numerical calculation. Calculated by comparing the results given the non-uniform quantization is superior to uniform quantization of the specific conclusions.