Accession Number : AD0294172

Title :   ALIASING ERRORS IN SAMPLED DATA SYSTEMS

Corporate Author : ADVISORY GROUP FOR AERONAUTICAL RESEARCH AND DEVELOPMENT PARIS (FRANCE)

Personal Author(s) : MALLINCKRODT, A.J.

Report Date : APR 1961

Pagination or Media Count : 12

Abstract : The principles of analysis of random time series are applied to the problem of quantitative estimation and minimization of aliasing errors. Message or signal functions are represented in terms of their frequency spectrum, and the result of sampling such message functions is analyzed in terms of the mean-square error of message recovery. Examples are treated that correspond to physically significant cases. It is shown that errors in practical systems can be such as to largely invalidate the resulting data. The analysis shows, for example, that, even with a message power spectrum which cuts off rapidly, a sampling-to-cut-off frequency ratio of at least ten-to-one is required for one per cent recovery error. The optimum recovery filter is shown to have not an infinite cut-off rate but one the same as that of the message power spectrum. The possibility of large sampling apertures is considered briefly and found to be a practical means of highly efficient pre-sampling filtering. The practical procedures for design of pre-sampling filters and measurement of the signal parameters required to set the sampling rate are discussed. (Author)

Descriptors :   *DATA PROCESSING, *RELIABILITY, ERRORS, FREQUENCY, RECOVERY, SAMPLING, SIGNALS.

Distribution Statement : APPROVED FOR PUBLIC RELEASE