Random No. generator
PaulHolland, there's no need to be rude.
You and I are saying basically the same things with different words. That paper says "we are considering shifts as different sequence". I say "same sequence with different starting point". Same thing.
I pointed out the zero problem, twice.
Of course the polynomial is important, hence my link to the Xilinx app note.
I do see one error in my earlier messages. Instead of saying "LFSR", I should have said "maximum-length LFSR". If that caused anyone trouble, I apologize. I believe the rest of my statements are accurate.