SS isn't redundant error correcting code. SS is accomplished by using
a pseudo-random bit sequence to chip the data stream into many smaller
bits, making the resulting channel stream look like Gaussian noise.
It's then de-spread at the receiving end and the baseband data is
recovered.
I think your best bet might be using COTS 802.11b (DSSS) hardware at
11Mbps on 3.4 GHz. Then there is the possibility of using 802.11a
hardware on that band. This would be a "simple" ;) matter of getting
hands on 802.11y hardware and tuning it down from 3.7 GHz. I don't
know what FCC would have to say about using 802.11a protocol on that
band. There used to be a prohibition in the rules about multiple
carrier signals in the same transmitter but I see no mention of it in
Part 97. I also see no mention of LFSR taps or length so it looks like
anything goes now. This sent me looking for my old Part 97 books but
it looks like I tossed them out.
FCC has already "clarified" their stand, "if it isn't specifically
prohibited, it's permitted", unless we tell you to stop. As for
clarification from ARRL or FCC, you don't even want to go there. FCC
has amended Part 97 to make it as broadly enabling as possible such
that if they don't want you to do something it's specifically
prohibited in Part 97 and if it's not prohibited it's permitted.
Unless they send you a cessation order after the fact the guiding
principle is to "just do it".
I see you doing a lot of complaining about what's NOT being done but
you don't seem to be doing something yourself. Grab that block of IP's
I allocated to you and start building. Then publish your results.
That's the basis and purpose of ham radio.