Re: [Discuss-gnuradio] Re: USRP2 802.11 BBN code on the TX side

[Brian Padalino]

On Mon, Dec 7, 2009 at 12:44 PM, George Nychis <address@hidden> wrote:
> I double checked the scrambler code and I really do not see anything wrong
> with it, it seems to meet the 802.11 spec for DSSS.  Then I found something
> in the spec that says that the transmitter need not start with the seed,
> since the receiver's descrambler is self-synchronizing.  So then I figured,
> maybe the transmitter just started with a different seed... so I generated
> 128 different signatures, for all 128 different possible scrambler seeds,
> and NONE correlate.
>
> Maybe it's using a different scrambling algorithm?  I'm running out of
> ideas.

It might transmit the opposite of what it is expecting to flush the
receiver scrambler and get it synchronized.  So if you need a long
length of 1's, maybe it starts sending a couple symbols worth of 0's
first - gets synchronized, then starts the sending of 1's.  Same with
the 0's for the preamble.

Maybe despread your symbols and run through the derandomizer, then
check to look at the pattern.  You should see your long string of 1's
or 0's (whatever you're looking for).  Then check to see what they
send beforehand?

Brian