| 摘要 |
For establishing word synchronization between a data bit stream of N-bit words and the timing of syndrome calculation, a count value is incremented in response to the bit timing of the words and generating a hunting pulse for every count of N bits and a load timing pulse at every count of C bits. Syndrome calculation is performed on the N-bit words in response to the hunting pulse to derive a syndrome when each of the N-bit words is out of synchronization with the timing of the calculation. The syndrome is latched for a word interval and a first enabling signal is generated in response to the latching of the syndrome. Syndrome generating words are counted to generate a second enabling signal when the count provides a valid indication that the word timing of the calculation is out of sync with the data bit stream. A copy of the hunting pulse is generated and delayed for a period of M bits when the first and second enabling signals are simultaneously present to cause syndrome calculation to be performed on the N-bit words, where the N and M integers are relatively prime and the integer C is smaller than integer M and greater than unity. The point of initialization is shifted for each syndrome calculation from one bit to another in search of the beginning of a word. Because of the "relatively prime" relationship of the integers N and M, all the bits of incombing data bit stream are examined in the search.
|
