Previously, new structures using complex arithmetic were proposed for the implementation of real coefficient IIR digital filters. The quantization of the real-equivalent of a complex coefficient in these filters violates an assumption usually adopted in the analysis of coefficient sensitivity, i.e., quantization errors are uncorrelated to each other. In the present paper, a simple expression is derived to measure the L/sub 2/ sensitivity, of structures using complex arithmetic. Then, an explicit expression of the L/sub 2/ sensitivity is obtained for second-order digital filters with minimum norm.
展开▼
机译:以前,提出了使用复数算法的新结构来实现实系数IIR数字滤波器。这些滤波器中的复数系数的实数的量化违反了通常在系数灵敏度分析中采用的假设,即,量化误差彼此不相关。在本文中,使用复杂的算术推导了一个简单的表达式来测量结构的L / sub 2 /灵敏度。然后,获得具有最小范数的二阶数字滤波器的L / sub 2 /灵敏度的明确表达。
展开▼