[time-nuts] PN sequence generation using GPS

[Bob Camp]

Hi

The key item here is that the system is going to work via amateur radio here in the US. The FCC only lets you use three very specific PN sequences. The three are called out explicitly in the rules. The requirement that they not be reset while transmitting except by feedback is also called out explicitly in the rules. So no fancy stuff with multiple spreaders and the like. 

The fact that they work out to primes is not important other than there's not much use trying to factor them. Without any common factors, and with 1 pps as your reset, you get a pretty specific trigger point for reset. It's long enough to not be of much use. 

There's a much easier way to do it and keep everybody in sync. GPS gives you time of day along with the 1 pps signal. You agree that at this or that time, everybody starts in sync. It's just time math to figure out where you should be at any 1 pps point. If you aren't where you are supposed to be, you take action. 

I suspect you could do all the checking on a 50 cent processor. Latch the data at 1 pps and read the time string. Do the math to check it. Fire the alarm flag it's not right. Not a lot to get done in a second. 

Bob

On Feb 16, 2011, at 7:22 PM, Magnus Danielson wrote:

> Bob,
> 
> On 02/16/2011 10:51 PM, Bob Camp wrote:
>> Hi
>> 
>> According to the rules, you are not allowed to reset shift register other
>> than by feedback during a transmission. In order for a reset to be
>> "transparent" it would have to always happen at the same time. Usually
>> that's when the register is full or empty.
>> 
>> A proper (MLS) PN sequence will be (2^m) -1 long. For the allowed ham
>> sequences that would be 2^7 -1 = 127, 2^13 -1 = 8191, or 2^19 -1 = 524287.
>> If they are not MLS then I'm off a bit. None of those divide into 10 MHz
>> very well at all, since they all are prime numbers.
> 
> It's actually not the prime-number aspects which is important, all Maximum Length Sequences (MLS) will not be prime-number lengths as not all Mersenne numbers is primes, and then non-prime polynomial lengths like 10 naturally breaks up (1023 = 31*33).
> 
> The key factor is that the sequences does not line due to lack of common factors. However, this deficiency can be overcome by shortening the sequence.
> 
>> If you use the shortest sequence, you would get to reset every 127 seconds
>> with a simple approach. You likely would do better to run a bit more math
>> and figure out the "right" sequence for the time you are at.
> 
> Probably more trouble than it is worth it.
> 
> Cheers,
> Magnus
> 
> _______________________________________________
> time-nuts mailing list -- time-nuts at febo.com
> To unsubscribe, go to https://www.febo.com/cgi-bin/mailman/listinfo/time-nuts
> and follow the instructions there.