Title :
Analysis of the recursive multiple window STFTs and spectrograms
Author :
Feng, Kai D. ; Amin, Moeness G. ; Tyler, Steve
Author_Institution :
Dept. of Electr. & Comput. Eng., Villanova Univ., PA, USA
Abstract :
A wide class of short time Fourier transforms (STFTs) or spectrograms can be generated by using an infinite-length analysis window corresponding to an impulse response of an IIR filter. The cascade form realization of the analysis window allows concurrent generation of several STFTs with different trade-offs between temporal and spectral resolutions. The number of STFTs or spectrograms associated with the analysis window is equal to the number of the W filter singularities. The temporal-spectral resolution diversity of the STFTs (spectrograms) generated by any given structure is a function of the filter poles and zeros as well as their order of cascade. This paper focuses on multiple pole filters and derives closed form expressions for their temporal resolution as a function of the pole position and multiplicity
Keywords :
Fourier transforms; IIR filters; cascade networks; filtering theory; poles and zeros; recursive filters; signal resolution; spectral analysis; statistical analysis; time-frequency analysis; IIR filter; cascade form analysis window; cascade order; closed form expressions; filter singularities; impulse response; infinite-length analysis window; linear time-frequency distributions; multiple pole filters; pole multiplicity; pole position; poles and zeros; recursive multiple window STFT; spectral resolution; spectrograms; temporal resolution; temporal-spectral resolution diversity; Bandwidth; Fourier transforms; IIR filters; Laboratories; Length measurement; Poles and zeros; Signal resolution; Spectrogram; Time frequency analysis; Time measurement;
Conference_Titel :
Time-Frequency and Time-Scale Analysis, 1994., Proceedings of the IEEE-SP International Symposium on
Conference_Location :
Philadelphia, PA
Print_ISBN :
0-7803-2127-8
DOI :
10.1109/TFSA.1994.467361
