Fuzzy-logic inference engines are in use in various disciplines such as con
trol systems, medicine, and the like. The use of optical tools to implement
such engines may improve the performance and the flexibility of inference
procedures. The optical processor works in a two-dimensional environment, w
hereas the inference engine might have to handle more than two independent
input channels. Here several approaches to generating the first, to our kno
wledge, N-dimensional optical fuzzy processor are addressed. The first appr
oach uses space multiplexing, the second approach uses polarization multipl
exing, and the third approach uses wavelength multiplexing to increase the
dimension of the processor. (C) 2000 Optical Society of America. OCIS codes
: 050.1970, 070.4560, 200.4660.