HYBRID ANALOG-DIGITAL TECHNIQUES AND RANDOM-PROCESS STUDIES

摘要

Iterative-differential-analyzer methods are presented which allow experimental studies of prediction networks to be performed without the use of delay lines. Two classes of problems are examined in this paper:n7. Prediction of stochastic signals with no noise present.n2. Prediction and filtering of deterministic signals masked by noise. Measurement techniques are developed for each case, and experimental verification is included.