V.F. Solomatin

The model of associative memory described, in which vectors components represented by sign-variable pseudorandom functions, which written in memory cells with superposition. Each association is in one cell written. This model is a transition from the neuronet model, proposed by the author earlier, and a "classic" associative storage device. The model has extended function ability and high noise-resistance. Possible application areas of devices with this structure include on-board systems, autonomous robots, pattern recognizing devices, electronic dictionaries, text translation systems, data analysis systems, etc.