The present invention is a method of determining codeword boundary without marker bits by receiving transmission bits; determining a dual code of a code used to generate the transmission bits; selecting a vector from the dual code; initializing n scoring variables; initializing i=1; initializing z=1; selecting n bits from the transmission bits starting at bit position i; performing a bit-wise AND operation on the vector and the n selected bits; if the result of the bit-wise AND operation contains an even number of ones then assigning a value of zero to the result, otherwise assigning a value of one to the result; setting S.sub.z equal to S.sub.z plus the result of the last step; if z is less than n, incrementing z and i each by 1 and returning to the seventh step, otherwise proceeding to the next step; if z=n, i<L, and it is desired to process additional transmission bits then incrementing i by 1, and returning to the sixth step, otherwise proceeding to the next step; identifying the scoring variable S.sub.z having the lowest score; and identifying the subscript z of the result of the last step as the codeword boundary.Method of identifying boundary of markerless codeword656074506/05/200321/07/200020031,019McCloskey; Joseph P.York; Eric V.US Patent and Trademark OfficeGoogle Patent Searchpatentimages.storage.googleapis.com/US6560745B1/US06560745-20030506-D00000.pngUnique