US 4,546,474 AGranted

Method of error correction

AssigneeSony Corporation

InventorsYoichiro Sako, Kentaro Odaka

Life of this patent

Estimating the term…

Loading patent…

Abstract

A method of receiving and decoding doubly-encoded, interleaved data occurring in blocks of information words and first and second check words to correct errors occurring due to transmission, includes adding error pointers to indicate uncorrectable errors, and thereafter checking the number of pointers in each block and protecting against miscorrection of errors. The data are decoded in a first decoder and the information words and first check words are corrected by syndromes using the second check words. A pointer is added to any words containing uncorrectable error. Then the words so decoded are deinterleaved and decoded in a second decoder, where the information words are corrected by syndromes using the first check words. Here the pointers show the position of any erroneous words to facilitate error correction. The number and location of erroneous words in each block, at the second decoder, is determined by using the error syndromes. If the erroneous word locations as determined by the syndromes is coincident with the locations indicated by the associated pointers, the erroneous words, up to the predetermined number, are corrected. Finally, uncorrected erroneous words are compensated by interpolation.

Claims14

Dependent claims

A method of decoding transmitted digital information signals to correct errors occurring therein as a result of transmission, wherein said information signals are received as blocks of interleaved digital data word signals and are applied to a first decoder as first error-correcting blocks comprised of a plurality of information word signals, a series of first check word signals associated with said plurality of information word signals, and a series of second check word signals associated with said plurality of information word signals, said information word signals and said first check word signals being interleaved, comprising the steps of:

decoding the received digital data word signals in said first decoder and correcting up to a predetermined number of the decoded digital information word signals and first check word signals by generating error syndromes using said second check word signals, the corrected and decoded digital information word signals forming a block of interleaved word signals;
tagging a pointer code signal to each of said information word signals of the block of interleaved word signals to represent whether an uncorrected error remains in such words, and, whenever there are more than said predetermined number of erroneous word signals in such block, giving all word signals thereof a pointer code signal indicating error;
delaying the tagged interleaved word signals of each such block of tagged interleaved word signals in a deinterleaving stage by respective different amounts so as to deinterleave such tagged interleaved word signals into a different arranging order, thereby providing a second error-correcting block;
and
decoding the digital information word signals of each such second error-correcting block in a second decoder by generating error syndromes using said first check word signals and correcting up to another predetermined number of erroneous word signals in such second error correcting blocks where the location of any erroneous word signal as calculated from said error syndromes is coincident with at least one of the erroneous word signals as indicated by the associated pointer code signals.

2 more elements

A method of error correction in transmitted data signals in which a first error correcting block is formed of one word signal included in each of a data series of plural channels in a first arranging state and a first check word signal for the one word signal;

said data series of plural channels and first check word series signals are delayed by different times at every channel to be converted to a second arranging state;
a second error correcting block is formed of one word signal included in each of said data series of plural channels and first check word series signals in the second arranging state and a second check word signal for the latter one word signal;
said first and second error correcting blocks are formed in accordance with error correcting code signals so that error syndromes are calculated and errors up to K words included in the same block can be corrected by obtaining error locations from the error syndromes;
a first decoding is carried out in a first decoder for the second error correcting block of data series of plural channels;
said data series of plural channels and first check word series signals in the second arranging state are delayed by different times at their respective channels to be converted to the first arranging state;
and thereafter a second decoding is carried out in a second decoder for the first error correcting block, said method of error correction comprising the steps of:
correcting errors up to a predetermined number of words in said second error correcting block during the first decoding;
tagging a pointer code signal indicating an error to every word signal in a block when error word signals exceeding the predetermined number are detected, thereby indicating that all word signals in the block are unreliable and are assumed to include an error;
and
calculating an error syndrome from a plurality of word signals included in the first error correcting block and hence correcting error word signals of a predetermined number by obtaining error locations based upon the error syndrome used in the second decoder.

7 more elements

Dependencies are inferred from the claim text.