Cable Forum - ntl's plans for the future?

Quote:

Originally Posted by ianathuth

Very interesting article.

Wonder what some will make of the statement:

Upstream
• The standard upstream spectrum is 5MHz to 65MHz using QPSK modulation in 3.2MHz channels
• Upstream capacity is 5.12 Mb/s per 2000 homes
• QAM 16 could increase this to 10.24 Mb/s to each home
Downstream
• The standard downstream spectrum is 85MHz to 750MHz using QAM 64 modulation in 8MHz channels
• 80 downstream channels are available
• Each downstream channel can support 38Mb/s of downstream data (Broadband and TV)
• Maximum downstream capacity is 80 x 38Mb/s = 3.04Gb/s delivered to each home
• QAM 256 modulation could increase this to 4Gb/s to each home

Nothing new in any of that, QAM256 was being talked about 4 years ago, the anmount of work to implement it was the killer. QAM16 on the return path is also a very difficult route.

QAM256 Downstream would be running at a reduced carrier to noise ratio if launched at the same levels as the existing QAM64, increasing the levels would put further strains on the plant. ie: composite second and third order distortions.
We all know you dont get something for nothing in this world, it would all be down to how much headroom in carrier to noise is currently available in the system and how higher in level they are able to drive the current plant in the field. There was much talk about headend combining when this was being discussed, there would be a fair amount of expenditure and downtime to recinfigure the combining and launch ampliers in many/most headends to accomodate this upgrade.

I'm just saying it's not like plug and play.

Moving form 3.2MHz QPSK to a QAM16 signal would reduce the signal to noise ratio on a system that is already marginal, much of the ingress is out of ntl's hands but network maintenance would have to be a lot tighter to ensure a good signal to noise ratio. launching with QAM16 upstream on some areas of the network in their current state would be a disaster, there is much cleaning up to do not helped by the decision to put a stop to general network maintenance in many if not most areas.

The HFC network is something that is gradually and continually degrading if not maintained. Managers who dont understand it are usually unaware and wonder why they have been wasting resources on it, then all of a sudden things get to a level where they see a big increase in customer calls.

To keep the extra system headroom for QAM16 they would need to keep the emphasis on network maintenance, I would go as far as to say they would need to increase the dwindling staff levels amongst the network techs to keep the quality of service at a level the customer would expect.

As I said, you dont get something for nothing. Higher symbol rate, more complex modulation scheme reduces the energy per bit over the wider bandwidth, going from the current QPSK to QAM16 would probably be on a par to reducing current signal to noise ration of somewhere near to 5dB off the top of my head.

The figures can all be worked out by ten log data rate x symbol rate x factor for modulation scheme/format. I think it's x2 for QPSK and x4 for QAM16.
again off the top of my head figures, I accept being corrected if I'm wrong.