Instruments and Systems: Monitoring, Control, and Diagnostics Annotation << Back Methodical apparatus of test-code error correction for random access memory A.A. PAVLOV, I.V. ROBERT, A.N. TSARKOV, P.A. PAVLOV, D.E. SOROKIN, A.A. BURMISTROV, D.D. IVANOV Proposed method of test-code error correction for random access memory consists in using the linear algebraic code for detection of occurring errors, and determining the configuration (erroneous digits) of single and double errors, correction of these errors is performed by value of the error syndrome and the analysis results of response reaction obtained by initiating the test impact. In contrast to known methods, proposed test-code method for errors detection and correction allows to: • correct 100% of single errors; • correct double errors in the information blocks; • have a minimum of information redundancy (two check digits, at any number of information bits); • ensure high reliability of operation (to detect 100% of possible errors, including errors transformable into allowed (error-free) code sets); • detect hidden errors; • minimize hardware expenses to build a decoding device (does not require storage for table of errors syndromes and uses the minimal table of corrections); • eliminates methodological error of decoding (typical for most decoding devices implementing the method of “maximum likelihood”: when errors with different multiplicity having the same syndrome belong to the same coset, are adjusted on the basis of the code distance of the selected linear code); • minimize the time required for error correction (requires initiation of a test impact that can only be initiated when errors are detected). Keywords: error correcting algebraic linear code, correcting cyclic code, single and double errors, hidden error, correction for hidden error, test vector of errors. Contacts: E-mail: Pavlov_iif@mail.ru Pp. 17-24.