数字信号处理定点除法精度算法与计算思维

数字信号处理中由于有限字长而存在精度问题,因此不可避免地存在误差,包括数据和系数的量化误差、计算上的舍入误差以及最终在算法和系统实现上产生的误差。计算思维近年来得到了广泛重视和大力发展,计算思维的培养需要多学科融合和长期实践训练过程。结合计算思维重要思想分析了定点DSP芯片上定点除法运算精度的算法,包括除法运算特点和定点除法程序设计,并以TI公司代表性芯片为例说明了操作特点,最后讨论了如何快速实现高精度定点除法。

Fixed-Point Division Precision Algorithm in Digital Signal Processing and Computational Thinking

In the digital signal processing, error is inevitable because of finite word length. It includes the quantitative error of the data and coefficients, the rounding error in the calculation, and eventually the errors generated in the algorithm and system implementations. In recent years, computational thinking has been widely valued and vigorously developed, and the cultivation of computational thinking needs a multidisciplinary integration and a long-term practice training process. Combining important ideas in computational thinking, this paper analyzes the fixed-point division operation, the fixed-point division program, the realization in representative DSP chip of TI, and finally how to quickly realize high-precision fixed-point division.