[opendtv] SFNs waste frequency spectrum
- From: "Manfredi, Albert E" <albert.e.manfredi@xxxxxxxxxx>
- To: "opendtv@xxxxxxxxxxxxx" <opendtv@xxxxxxxxxxxxx>
- Date: Thu, 8 Jul 2010 18:18:19 -0500
Al Limberg wrote:
> SFNs are wasteful of spectrum, severely reducing the amount of
> payload that can be transmitted compared to using repeaters on
> different channels, supposing one wishes to avoid spectrum nulls
> in regions where coverage areas overlap. This is very evident
> with 8VSB, where spectrum nulls are best avoided by staggering the
> transmission times of the various transmitters.
I take it, you're talking about synchronizing the transmitters. Not introducing
a guard interval in 8-VSB, which would provide little relief given how fast the
symbols are transmitted.
> But it is also the case with COFDM, where the coding to overcome
> spectrum nulls is the mechanism that reduces the amount of payload
> that can be transmitted.
Yes indeed. In the Italian SFNs, for example, they use GI of 1/4. Very
wasteful. The Germans use 1/8. The amount of payload lost is expressed by the
GI, all else equal. Compared with no GI at all, 1/4 GI throws out 25 percent of
the channel capacity. Not negligible.
But with DVB-T2, the symbols are even slower in 16K and 32K modes, so the price
you pay is not so high. Just as a rough estimate, with 8K mode and 1/8 GI in
DVB-T, the towers can be about 11-12 miles apart safely, with a passive SFN
(round trip time between towers covered by the GI). So if 32K mode creates
symbols that are 4X slower that 8K, the distance between towers can be extended
to maybe 45 miles. Which makes the regional SFN a little more credible.
Certainly, a Wash/Balt region could be covered well enough, since the existing
towers are closer together than that already. Between Balt and Phila you would
need some intervening towers, but not a ridiculous number as you would need in
DVB-T (and still cover the entire area continuously).
> A regional broadcast system would preferably employ frequency
> diversity, so signals from two RF channels could be combined during
> parallel digital processing of the two channels in the receiver.
> This would provide better reception in regions where coverage areas
> overlap, without requiring spatial-diversity antennas.
Not sure I understand this part. I get the inverse multiplexing, but I don't
see where using two frequencies helps in area of overlap. If one or the other
frequency gets messed up, haven't you lost the signal anyway?
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.
Other related posts:
