Thirty-two adjacent pairs of lattice points in the cruciform 512QAM point lattice must have 2-bit rather than single-bit transitions between them. I described one such point lattice in published patent application US-2013-0028271-A1. That application described turbo coding, rather than LDPC coding, replacing the convolutional coding of DVB-T. Since there are nearly two thousand adjacent pairs of lattice points in the point lattice for 512QAM, and 9 bits associated with each lattice point, this is pretty close to perfect Gray mapping. However, Burt is right. There will be some degradation in AWGN performance owing to the double-bit transitions between adjacent lattice points. Allen ----- Original Message ----- From: "Albert Manfredi" <albert.e.manfredi@xxxxxxxxxx> To: <opendtv@xxxxxxxxxxxxx> Sent: Saturday, March 30, 2013 6:15 PM Subject: [opendtv] Re: With CFP, Air Test, ATSC 3.0 Off And Running | TVNewsCheck.com > Allen Limberg wrote: > > > One way to reduce PAPR is to use 128QAM and 512QAM with > > cruciform symbol constellations rather than 64QAM and > > 256QAM with square symbol constellations. The 128QAM and > > 512QAM can be used with single-time retransmission to get > > back AWGN performance. Single-time retransmission (used > > by DVB-T2 in their rotated constellation option) > > overcomes burst noise and long drop-outs in received > > signal strength. If circular FFT is rotated one half > > revolution as between two transmissions, deep selective > > fading is overcome, too. > > Thought-provoking as always, Al. So as always, everything comes at a price. I looked up these cross constellations, and it seems that because you cannot achieve a case where each adjacent constellation point differs always by only one bit, from any given point on the constellation, as you can achieve with square constellations, this increases the bit error rate. > > Got to do more digging into these trades. But it does appear like Mark Aitken can indeed fine-tune what DVB-T2 has standardized, to tweak the coverage. > > Bert > > ---------------------------------------------------------------------- > You can UNSUBSCRIBE from the OpenDTV list in two ways: > > - Using the UNSUBSCRIBE command in your user configuration settings at FreeLists.org > > - By sending a message to: opendtv-request@xxxxxxxxxxxxx with the word unsubscribe in the subject line. > ---------------------------------------------------------------------- You can UNSUBSCRIBE from the OpenDTV list in two ways: - Using the UNSUBSCRIBE command in your user configuration settings at FreeLists.org - By sending a message to: opendtv-request@xxxxxxxxxxxxx with the word unsubscribe in the subject line.