Title :
A fuzzy associative memory module and its application to signal processing
Author :
Nodes, T.A. ; Smith, J.L. ; Hecht-Nielsen, R.
Author_Institution :
TRW-MED, San Diego, CA, USA
Abstract :
A fuzzy associative memory or FAM is a device in which data in memory is located and manipulated based on the value of the data itself. In the FAM, this processing is based on ranges of values (fuzzy) not specific points. This allows such a device a great deal more applicability to real world problems. This paper describes such a device currently under development at TRW. This device is intended as a high speed association coprocessor for signal and data processors. In addition, the FAM´s potential for supplying practical solutions to previously difficult multivariate detection and estimation problems is investigated and some analytic tools presented.
Keywords :
Associative memory; Coprocessors; Expert systems; Floating-point arithmetic; Hardware; Pattern recognition; Signal analysis; Signal processing; Throughput; Very large scale integration;
Conference_Titel :
Acoustics, Speech, and Signal Processing, IEEE International Conference on ICASSP '85.
DOI :
10.1109/ICASSP.1985.1168070