Title :
Parallel time-frequency analysis
Author_Institution :
Rockwell Int., Thousand Oaks, CA, USA
Abstract :
Conventional time-frequency analysis methods are generally too computationally intensive for real-time implementation. A relatively new method based on a dynamic state-space model is inherently a real-time method and can be implemented in parallel processing. It has O(n) computational complexity per signal sample, where n is the number of frequencies in the analysis. Its offered concurrency allows only sublinear speedup with parallel digital implementations, but linear speedup with a partially analog implementation
Keywords :
computational complexity; computerised signal processing; parallel processing; real-time systems; state-space methods; computational complexity; dynamic state-space model; linear speedup; parallel digital implementations; parallel processing; parallel time frequency analysis; partially analog implementation; real-time method; signal sample; Computational complexity; Concurrent computing; Damping; Discrete Fourier transforms; Parallel processing; Signal analysis; Signal processing; Sonar; Speech; Time frequency analysis;
Conference_Titel :
Acoustics, Speech, and Signal Processing, 1989. ICASSP-89., 1989 International Conference on
Conference_Location :
Glasgow
DOI :
10.1109/ICASSP.1989.266909
