A coupled multichannel filter bank and sniffer spectrum analyzer

Abstract

The FFT, the efficient algorithm for implementing the DFT, enjoys great acceptance as the signal processing tool for spectrum analysis, for channelized receivers, and for fast convolution. In the first applications, spectrum analysis, the FFT is supported by a set of weights, the window, applied to data multiplicatively. In the second application the FFT is supported by a set of weights, the filter, applied to data convolutionally. Both operations accomplish the same task; that of spectral decomposition with controlled spectral response. In reality, the two operations are identically the same since the sliding windowed FFT is in fact a particular implementation of a resampling filter bank. Since the two processes are the same, when a system includes both a channelizer and a spectrum analyzer that steers the channelizer to spectral areas of interest the two can be combined or coupled to share their computational burden. In this paper we illustrate the benefit of this merged option.