FPGA-based quadruple wavelet multi-channel signal processing method

Abstract

The invention discloses an FPGA (field programmable gate array)-based quadruple wavelet multi-channel signal processing method, which comprises the following steps of obtaining a quadruple wavelet coefficient, wherein the quadruple wavelet coefficient comprises a quadruple wavelet decomposition coefficient and a quadruple wavelet reconstruction coefficient; optimizing the quaternion wavelet coefficient to reduce an error of the quaternion wavelet reconstruction; converting the optimized quaternion wavelet coefficient into a structure combining integer multiplication with shift operation; carrying out parallel operation on a signal to be processed by using an FPGA platform in combination with the quaternion wavelet coefficient and according to the structure combining integer multiplicationwith shift operation, wherein the parallel operation includes the quaternion wavelet decomposition, extraction interpolation and quaternion wavelet reconstruction processing; and carrying out zero setting, compression, threshold setting and cascade filtering in a quaternion wavelet reconstruction process according to the actual conditions. According to the method, the characteristic processing ofmultiple frequency domains can be carried out on the digital signal, the decomposition level of the quadruple wavelet filter can be improved through actual conditions to meet different resolution requirements, the operation complexity is low, the operation speed is fast, and the hardware operation resources are saved.

Description

technical field [0001] The invention belongs to the field of signal processing, in particular to an FPGA-based four-entry wavelet multi-channel signal processing method. Background technique [0002] The core of signal processing is digital filtering. Digital filters can be divided into two parts: classic filters and modern filters. Classical filters assume that the spectrum of noise and signal belong to different frequency ranges, and noise removal can be achieved by setting the passband and stopband in the frequency domain. Commonly used high-pass, low-pass, band-pass, and band-stop filters are all classical filters. Because the classical filtering system mainly filters out the fixed frequency band signals that reflect the characteristics of the electronic system and the real signals are often distributed in this frequency band, so for the weak signal processing required by high resolution, the classical filtering cannot meet the requirements, and the useful signal and Mo...

AI Technical Summary

Problems solved by technology

Binary wavelets cannot take into account symmetry, linear phase and tight support at the same time (except Haar wavelet, but the Haar wavelet function loses smoothness)

However, there are problems in the application of multi-entry wavelet to multi-channel filters: it is difficult to obtain the coefficients of multi-entry wavelet, and the coefficients except the coefficients in the Haar array are irrational numbers; but the lack of smoothness of the Haar array makes it difficult to complete real-time The task of signal processing; and the realization of irrational number multiplication on the FPGA based on fixed-point arithmetic has a word length effect, and complete reconstruction cannot be guaranteed

Images