[LINK] Net neutrality and bandwidth caps.

[Tom Koltai]

Heh heh heh.

Im so glad someone else took this one up.

> -----Original Message-----
> From: link-bounces at mailman1.anu.edu.au 
> [mailto:link-bounces at mailman1.anu.edu.au] On Behalf Of Robin Whittle
> Sent: Wednesday, 6 May 2009 8:44 PM
> To: Link
> Cc: Kim Holburn
> Subject: Re: [LINK] Net neutrality and bandwidth caps.
> 
> 
> This is completely untrue:
> 
> >> Bandwidth costs
> >> are almost zero, and given how things are set up, there is no way a
> >> single person or even a small number can max out the 
> bandwidth of a  
> >> cable loop.
> 
> HFC cable has limited downstream bandwidth which must be 
> shared, depending on how many 6 to 8MHz RF channels are 
> devoted to cable modems.  These are about 4 bits per Hz, if I 
> remember correctly.
> 
> The upstream link is far worse, with upstream only between 30 
> and 45 or so, maybe a little higher.  I recall less than 1 
> bit per Hz there.
> 
> Since maybe as much as 50% of traffic is peer-to-peer:
> 

Actually - 57% according to whirlpool.

> http://www.ipoque.com/resources/internet-studies/internet-stud
> y-2008_2009
> 
> with typically symmetrical upstream and downstream, it is 
> clear that this is not going to work out well on a cable modem system.
> 
> A cable system may server 1000 homes or so.  If 300 of them 
> have cable modems, operating at the same time - especially 
> with P2P file sharing programs which run 24 hours a day - it 
> is easy to see that this sort of usage does come up against 
> hardware limits.
> 
> It would be possible to break the system into 10 with 100 
> homes each.  That would help a lot, but it involves taking 
> fibre deeper into the access network, and more expense with 
> opto-electronics there and in the head-end.
> 
> Upstream is less efficient per Hz due to mixing of noise from 
> all sources, including the cumulative noise of the upstream 
> amplifiers in series, and multiple tributaries of these 
> feeding the electro-optical unit which converts the upstream 
> frequencies to an RF modulated optical signal.  Also, there 
> is the need for multiple cable modems to take it in turns to 
> transmit, so even if half the frequency range (30 to 750MHz 
> is the typical range) was used for upstream, there would 
> still be about 1/4 the upstream bandwidth compared to downstream.
> 
> For historical reasons, including compatibility with existing 
> cable TV system which I understand had FM channels (or is it 
> to avoid cable modems transmitting at FM frequencies, and due 
> to the need for a substantial guard band between upstream and 
> downstream (analogue filters in the amplifiers), HFC is bound 
> to be suffering from limited upstream capacity.  Also, I 
> think both HFC and ADSL were designed with the fantasy of 
> customers being good little consumers, sucking lots of "content".
> 
> I am not suggesting that there are no problems with bandwidth 
> caps etc.  Just that if there is going to be an argument 
> against them, it needs to be based on the realities of the 
> HFC and DSL systems, and the realities of the upstream costs. 
>  The sentence I objected too makes me think of a spoilt child 
> insisting that everything they want is free and should be 
> made available to them without any fuss.
> 
>   - Robin
> 

Hear hear.

Tom




_______________________________________
No viruses found in this outgoing message
Scanned by iolo AntiVirus 1.5.6.4
http://www.iolo.com