Methods and formulas for computer simulation studies of the performance of error-correcting codes for the North American Basic Teletext System (NABTS)

The rationale, methods and formulas used in computer simulation studies of the error performance of the various constituent error-correcting codes designed for the North American Basic Teletext System are described for the case when the input error sequences have an arbitrary distribution. Expressions for the overall packet rejection rate, R, and the number of picture-description instruction (PDI) errors are derived using a classification of the undetected errors in the extended Hamming (8,4) encoded sections into those giving rise to delay, and those giving rise to corruption. Realistic bounds for acceptable packet rejection rates and number of PDI errors per record are derived in an attempt to relate the performance of each code to a user´s perception of the system performance. To validate the computer simulation program a comparison of the simple product code, code C, and the bundle code is made by generating independent errors at a desired probability of error, using these error sequences as inputs to the coding analysis program, and then comparing the performances thus obtained with the performances predicted by the theory