High speed real-time design and implementation of Cauer-type Jaumann digital filters

Presents an investigation of the design and real-time implementation of lossless discrete-integrator Jaumann digital filters. The salient feature of these digital filters is that they incorporate a minimal number of digital multipliers and have low sensitivity to multiplier coefficient and data wordlengths. Fast operational speed is achieved by using a Cauer-type structure which inherently is of the order of the transfer function to be realized. The investigation involves a fixed-point bit-serial implementation as well as a real-time floating-point digital signal processor (DSP) implementation using AT&T DSP32 and TI TMS320C30. Examples are given to illustrate the results. It is shown that it is possible to achieve a sample frequency of around 1 MHz in the case of a fixed-point implementation, and sample frequencies of about 180 kHz in the case of a floating-point DSP processor implementation.<
>