Suppose I have selected an appropriate CRC polynomial and know the content and size of the data frame I want to protect ... How do I choose the optimal initial value?
It is known that CRCs with an initial value of 0 have a blind spot for messages with leading zeros. For this reason, many articles and websites mention / recommend an initial value of all-ones, as this fixes the blind spot.
My question: Does all-ones have any disadvantages? Can there be a better initial value than all-ones if the specific known data frame is taken into account?