[opendtv] Re: 44 Mbps in 8Mhz for a UK-wide SFN ?
- From: "Bob Miller" <robmxa@xxxxxxxxx>
- To: opendtv@xxxxxxxxxxxxx
- Date: Thu, 20 Mar 2008 22:35:28 -0400
On Thu, Mar 20, 2008 at 9:12 PM, Albert Manfredi <
albert.e.manfredi@xxxxxxxxxx> wrote:
>
> Olivier Houot wrote:
>
> > This is suggested in this interesting test report :
> >
> http://www.bbc.co.uk/rd/pubs/whp/whp-pdf-files/WHP157.pdf
> >
> > It seems using two orthogonal polarizations works well,
> > after all, when coupled with the MIMO concept.
>
> Theu use horizontal and vertical paths as the two propagation paths,
> instead of relying on reflections. I think this is a good idea, for one-way
> broadcast. MIMO otherwise only works well when you have two-way links, so
> that conditions can be assessed in real time between two positions. And they
> are figuring about 10 dB of separation between the two, it looks like.
>
> > One other interesting item they mention is that DVB-T2 is
> > likely to add a 32 K carriers mode, which would make
> > nation-wide SFNs easier.
>
> My take is that this is a necessary measure, to make a robust SFN, which
> is needed to compensate for the probably relatively short range per tower.
>
> The basic signal seems to be a 1/8 GI, 1/32 FEC, 64-QAM conventional
> COFDM, which provides 22 Mb/s in the 8 MHz channel. So it looks like two of
> these 22 Mb/s paths, one H and one V, with attenuated cross terms. So this
> drove the system to planning factors more conservative than standard COFDM.
> They called for very low antenna feeder loss and equal signal density,
> compared with the more conventional alternative, and greater C/N margin. At
> Euro power levels, this means short range.
>
> Perhaps that was a good motivation for a SFN approach, which in turn drove
> the need for more subcarriers. By adding a gozillion subcarriers, they
> extended the symbol duration and GI enough to make for a safer large area
> SFN.
>
> > I am not sure 44 Mbps is the limit, though. As a minimal,
> > DVB-T2 aims for a 30% increase in datarate, which means
> > 31 Mbps could be achieved with the same reliability as 24
> > Mbps nowadays. So it would seem possible to reach
> > something like 60 Mbps with 2-antennas MIMO.
>
> Maybe so, if they introduced something like the training sequence of the
> Chinese system. But there's no indication of that, so the system relies on
> active carriers just like the current COFDM, as far as I can tell.
>
> The problem I see with this is infrastructure costs. Not the sort of thing
> US broadcasters would happily entertain on their own nickel, I suspect.
>
> Bert
Not current broadcasters with current broadcast mindset. The cost of
infrastructure is low compared to value if correctly valued. And who knows
how low infrastructure cost might be if someone were to use non conventional
thinking there also.
Bob Miller
Other related posts:
