Associative network and signal handling element therefor for processing data

Abstract

A signal processing network circuit which may be used, for example, as a filter or an associative memory has elements (101, 102, 103, 104) each having a respective input conductor (85, 86, 87, 88) and a respective output conductor (81, 82, 83, 84) a teach cell (T) and a recall cell (R). Each input conductor (85, 86, 87 or 88) has a number of memory cells (M) serving as signal coupling devices between output conductors of other generators and the input conductor. In a teach phase, paths through the memory cells (M) are interrupted whenever the output conductor is high and the input conductor is low. Each teach cell causes the input conductor of the respective generator to be low if the output conductor is high. In a recall phase, the recall cells (R) receive input signals from their input conductors in dependence upon the number of output conductors coupled thereto through memory cells (M). If the input signals received exceed a threshold, the recall cell (R) sets the output conductor low, and vice versa. The input signals are summed and stored at the recall cells (R), but decay with time so that after a time a fixed pattern of high and low states appears on the output conductors, and can be sensed by a sensing circuit (90) which enables an output buffer (89).

Dependencies are inferred from the claim text.