Switching current Gauss low-pass filter

Abstract

The invention relates to a switching current Gauss low-pass filter which is comprehensively implemented by using the central limit theorem to obtain a transfer function of a Gauss low-pass filter and adopting a switching current bilinear integrator cascade structure with a plurality of negative feedback branches. The switching current Gauss low-pass filter comprises a current mirror circuit, a plurality of first switching current bilinear integrators and a second switching current bilinear integrator, which are in cascade connection sequentially. The switching current Gauss low-pass filter has the following beneficial effects: the filter is constructed by adopting a switching current bilinear integrator cascade structure with a plurality of negative feedback branches, so that the sensitivity of the filter is effectively reduced, and the actual frequency response characteristic of the circuit is enabled to be close to the ideal frequency response characteristic; and the switching current Gauss low-pass filter is large in dynamic range, high in approximation accuracy and simple in design process, needs no analog-to-digital converter, and is applicable to low-voltage and low-power large-scale integration.

Description

technical field [0001] The invention relates to a switch current Gaussian low-pass filter, which belongs to the technical field of signal processing. Background technique [0002] Due to its minimum time-bandwidth product, Gaussian filters have been widely used in communication systems, image processing, computer vision and other fields. Gaussian filters are mainly divided into Gaussian low-pass filters, Gaussian band-pass filters and Gaussian high-pass filters. Among them, the Gaussian band-pass and high-pass filters can be obtained from the Gaussian low-pass filter through frequency conversion, so the design of the Gaussian low-pass filter has become the focus of attention. At present, the implementation circuits of Gaussian low-pass filters are mainly divided into two types: digital type and analog type. Digital Gaussian low-pass filters have the disadvantage of high power consumption due to the need for an analog-to-digital converter. In contrast, the analog Gaussian ...

AI Technical Summary

Problems solved by technology

Although the feasibility has been verified by examples, there are still many shortcomings in this implementation method

In terms of rational fractional approximation of Gaussian functions, the solution formula of the parameter τ in the approximation fraction is not accurate enough, so the approximation accuracy of the approximation fraction does not increase significantly with the increase of the order n

In terms of the realization of the filter structure, the cascaded structure has high sensitivity to the parameter values ​​of the components, resulting in the actual frequency response characteristics of the filter being not ideal.

Images