Title :
Distributed filtering without knowledge of noise distribution
Author_Institution :
Dept. of Electr. Eng., Wright State Univ., Dayton, OH, USA
Abstract :
An algorithm of distributed filtering using set models with confidence values is derived. No knowledge of the noise distribution is needed. The only information required is the set with confidence values from which the modeling and measurement errors and the initial values are obtained. Therefore, the algorithm has great potential for real-world applications
Keywords :
filtering and prediction theory; set theory; algorithm; confidence values; distributed filtering; measurement errors; modeling errors; set models; Distributed algorithms; Filtering algorithms; Filters; Measurement errors; Smoothing methods; Stochastic processes;
Conference_Titel :
Acoustics, Speech, and Signal Processing, 1992. ICASSP-92., 1992 IEEE International Conference on
Conference_Location :
San Francisco, CA
Print_ISBN :
0-7803-0532-9
DOI :
10.1109/ICASSP.1992.226334
