[opendtv] Re: COFDM and equalization
- From: "Manfredi, Albert E" <albert.e.manfredi@xxxxxxxxxx>
- To: <opendtv@xxxxxxxxxxxxx>
- Date: Wed, 21 Apr 2004 12:05:32 -0400
Bob, the difference there is simply caused by the trellis code
used. Or, in other words, if you use 2/3 coding like ATSC, you
are dedicating 1/3 of the bit rate to convolutional FEC.
As we discussed some time ago, in round numbers, both
64-QAM and 8-VSB give you exactly the same bit rate in a
given channel. In the 6 MHz channel, round numbers, 30
Mb/s.
With 8-VSB, you reduce that to 20 Mb/s when you add trellis
coding, then a little more for Reed Solomon, and a little more
for other overhead, leaving you 19.39 Mb/s for payload.
With COFDM, you lose some to GI but you *may* gain it back
by applying less of a convolutional FEC. So your 19.76 Mb/s
example applies only a 3/4 FEC, and a 1/16 GI, to 64-QAM.
Now to make that more like 8-VSB, apply 2/3 FEC and the
minimum 1/32 GI, and you will get 18.096 Mb/s with
COFDM, compared with 19.39 in 8-VSB.
So that's the bit rate comparison.
To compare robustness apples with apples, it makes sense to
tune both systems for the same bit rate, or as close as
possible. So IMO they did the right thing to tune COFDM
to 19.76 Mb/s for that test.
Bert
> -----Original Message-----
> From: Bob Miller [mailto:bob@xxxxxxxxxx]
> Sent: Wednesday, April 21, 2004 10:25 AM
> To: opendtv@xxxxxxxxxxxxx
> Subject: [opendtv] Re: COFDM and equalization
>=20
>=20
> COFDM has a penalty, a cost in data rate that comes with the=20
> use of GI.=20
> Understood.
>=20
> So what is the penalty that 8-VSB suffers that limits it to=20
> 19.34 Mbps?=20
> Or do they just throw away bits for no reason?
>=20
> The question is asked because in the hearings in 2000 Sinclair was=20
> broadcasting at 19.76 Mbps and was able to receive on a simple $2=20
> antenna while walking around the hearing room in a mobile or at least=20
> pedestrian mode.
>=20
> They were using a GI, they were taking a hit, yet they were=20
> at a higher=20
> datarate than 8-VSB and 8-VSB had to use a special=20
> directional antenna=20
> in a very precise location, that is they got lucky. I know I was=20
> watching them set up the night before and they were sweating bullets.
>=20
> So why was (and is) COFDM capable of a higher data rate with=20
> a GI than=20
> 8-VSB? Where is the penalty or more precise in what way is=20
> 8-VSB being=20
> penalized down to 19.34 Mbps?
>=20
> Bob Miller
----------------------------------------------------------------------
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.
Other related posts:
