Accession Number : AD0663738

Title :   SUB-SEQUENCE CORRELATION ANALYSIS.

Descriptive Note : Final rept. 20 Dec 66-19 Oct 67,

Corporate Author : CHESAPEAKE SYSTEMS CORP COCKEYSVILLE MD

Personal Author(s) : Cooper,A. Brinton ; Lord,Peter H.

Report Date : NOV 1967

Pagination or Media Count : 97

Abstract : One method of synchronizing a communication system uses the cross correlation between a maximal length pseudorandom, binary sequence and a portion of that sequence. Using only part of the sequence causes the correlation function to have peaks when the sequences are not aligned, and thus may cause a threshold decision scheme to synchronize incorrectly. An analysis was performed on the synchronization process by which a maximum length, linear, pseduorandom binary sequence is cross correlated with a very short sample of itself until correlation occurs. The object of the analysis was to determine which of many available pseudorandom sequences of a given length would be least likely to cause abnormally large correlator outputs in the un-correlated state. A second and equally important goal was the specification of the optimum decision threshold to be used to achieve a minimum probability of false indication of synchronism while maintaining a uniformly low probability of falsely dismissing the correct synchronous state. The investigation centered upon several theoretically derived hypotheses concerning the sequence generating functions and upon the experimental verification of these hypotheses. Data accumulated during the latter effort was used to calculate the optimum thresholds for the noise-free and the noise-present cases. (Author)

Descriptors :   (*CORRELATION TECHNIQUES, SEQUENCES(MATHEMATICS)), (*INFORMATION THEORY, MATHEMATICAL ANALYSIS), CODING, SAMPLING, POLYNOMIALS, TRANSCENDENTAL FUNCTIONS, FEEDBACK, OPTIMIZATION, THEOREMS

Subject Categories : Operations Research
      Cybernetics

Distribution Statement : APPROVED FOR PUBLIC RELEASE