Title :
New abrupt change detector using Walsh functions
Author :
Robert, Thierry ; Mailhes, Corinne
Author_Institution :
GAPSE, ENSEEIHT, Toulouse, France
Abstract :
This paper deals with parametric model abrupt change detection. Within this field, a time varying autoregressive (TAR) modeling is presented, using Walsh functions as basis functions. Then, a new abrupt change detector based on this modeling is derived. An associated deflation algorithm leads to a complete detection tool with a maximum error of about 2 or 4 points around the estimated value
Keywords :
Walsh functions; autoregressive processes; parameter estimation; signal detection; signal representation; spectral analysis; time-frequency analysis; time-varying systems; Walsh functions; abrupt change detector; basis functions; deflation algorithm; maximum error; parametric model; signal representation; time varying autoregressive modeling; time-frequency representations; Change detection algorithms; Detectors; Difference equations; Parametric statistics; Polynomials; Signal processing; Spectral analysis; Time frequency analysis; Vectors; White noise;
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.1994.467255
