Loading ...
Sorry, an error occurred while loading the content.

up/down bandwidth price discrepancy

Expand Messages
  • Sampo Syreeni
    Hi. I ve for the longest time wondered why the upload and download bandwidths are priced so disproportionately, and why that discrepancy has found its way even
    Message 1 of 4 , Jun 28, 2010
    • 0 Attachment
      Hi. I've for the longest time wondered why the upload and download
      bandwidths are priced so disproportionately, and why that discrepancy
      has found its way even into rigidly asymmetric standards like docsis.
      I've seen a million proximal explanations for that after having asked,
      but not one of them has seemed like an ultimate one. So as my first
      question, does anybody have any citations into publicly available
      research which tries to explain the phenomenon on bare, fundamental,
      economic grounds?

      My second question is then what sparked the first one. Yesterday I
      caught wind of a very disturbing news item, which suggested Australia
      might be going along with Chinese style, fully firewalled and blacklist
      enabled internet connectivity. To fight the Four Horsemen, basically. So
      immediately I thought, better start spreading Tor propaganda among my
      friends. One of them then brought up the fact that since all Tor traffic
      has to go through both up and down bands, that makes the system really
      slow, *even* after we have lots of users on board to run the mixnet.

      Still, at the very same time, I was delivering content to an online
      acquaintance of mine in Switzerland at 2.5MBps. That service costs me
      something like 30 euros per annum. That sort of thing also seems to be
      the norm: you can get colo upload speeds for a fraction of the cost of
      getting the same domestically. And yet your shell provider probably
      won't let you put up a permanent piece of any P2P architecture.

      Thus the question is, is there any research into protocols which could
      amplify Tor-like realtime onion routing/mixnet protocols so that they
      might use ordinary non-rerouted web servers for the actual delivery of
      content? Preferably with Torrent-like game theoretical strategy
      thinking, deniability of active ill will while using them (perhaps even
      full repudiation after the fact like in certain online chat protocols),
      or even anonymous offline clearing of bits for bits/money, as in
      bandwidth-as-anonymous-cash?

      To me at least that sort of thing could blow most of this asymmetricity
      right of the water, and enable the upload/download tradeoff to be placed
      arbitrarily within the collective cloud, wherever it's technologically
      most efficient. And to be blunt, if somebody pulled that off, the early
      cypherpunk dreams of what anonymous cash can do to the political sphere
      wouldn't then be too far behind... ;)
      --
      Sampo Syreeni, aka decoy - decoy@..., http://decoy.iki.fi/front
      +358-50-5756111, 025E D175 ABE5 027C 9494 EEB0 E090 8BA9 0509 85C2
    • Wes Felter
      ... DOCSIS has a good reason for being asymmetric; apparently the upstream channel has to be on a low frequency (
      Message 2 of 4 , Jun 28, 2010
      • 0 Attachment
        On 6/28/10 5:15 PM, Sampo Syreeni wrote:
        > Hi. I've for the longest time wondered why the upload and download
        > bandwidths are priced so disproportionately, and why that discrepancy
        > has found its way even into rigidly asymmetric standards like docsis.

        DOCSIS has a good reason for being asymmetric; apparently the upstream
        channel has to be on a low frequency (<42MHz) due to the design of the
        cable plant, and there isn't room for many upstream channels.

        In DSL there is a tradeoff between upstream and downstream bandwidth due
        to the use of FDD in the physical layer, so in the beginning they
        optimized it for downloading video and never revisited that decision.
        (And do you *really* want 2.3/2.3 instead of 8/1?)

        I don't know why PON is asymmetric other than reducing ONU cost.

        Wes Felter
      • Sampo Syreeni
        ... That is a perfect example of what I called a proximate cause, which fails to be an ultimate one: why precisely can t my side of the cable be just as good
        Message 3 of 4 , Jun 28, 2010
        • 0 Attachment
          On 2010-06-28, Wes Felter wrote:

          > DOCSIS has a good reason for being asymmetric; apparently the upstream
          > channel has to be on a low frequency (<42MHz) due to the design of the
          > cable plant, and there isn't room for many upstream channels.

          That is a perfect example of what I called a proximate cause, which
          fails to be an ultimate one: why precisely can't my side of the cable be
          just as good as the opposite side? Put out and have received just as
          many bits as it does? There's bound to be an economic reason behind
          that, somewhere, but I'm not seeing it. E.g. it might be that most
          people just do not want to pay for that level of equipment, true. Yet,
          some of us might. Why can't we just pay for it, have the standard allow
          it, and then e.g. have a gigabit per second worth of
          competitively-priced-upload band, traded in for a mere 10mbps download
          one? As an ordinary consumer?

          > In DSL there is a tradeoff between upstream and downstream bandwidth
          > due to the use of FDD in the physical layer, so in the beginning they
          > optimized it for downloading video and never revisited that decision.

          Again you beg the question: if FDD (whatever it is; FDM I know but not
          FDD, yet) is the bottleneck, which economic incentive ultimately caused
          it to be engineered into the protocol? To make it really simple and
          twisted...why can't I, personally, invert the situation and buy a set of
          lines where *I* have the central equipment and the "usual provider" has
          usual peripherals? At the going price? On a competitive market that
          should be a no-brainer; competitive markets are fully inversible because
          there is not profit, or entry or transaction cost.

          That sort of asymmetry is much deeper than what you say, and I've never
          seen it explained in full.

          > (And do you *really* want 2.3/2.3 instead of 8/1?)

          At times, yes: when my Torrent partners go into the tit-for-tat mode.
          Plus, if you'd looked at the rest of my post, you might see wider
          applications for that and/or incentives to go there. ;)

          > I don't know why PON is asymmetric other than reducing ONU cost.

          PON? ONU? Do tell.
          --
          Sampo Syreeni, aka decoy - decoy@..., http://decoy.iki.fi/front
          +358-50-5756111, 025E D175 ABE5 027C 9494 EEB0 E090 8BA9 0509 85C2
        • Rich Persaud
          ... From http://www.hyperorg.com/blogger/2009/10/03/why-up-is-slower-than-down : .. we download faster than we upload because of economics, not physics. The
          Message 4 of 4 , Jun 28, 2010
          • 0 Attachment
            On 6/28/2010 9:49 PM, Sampo Syreeni wrote:
             

            That sort of asymmetry is much deeper than what you say, and I've never
            seen it explained in full.


            From http://www.hyperorg.com/blogger/2009/10/03/why-up-is-slower-than-down :

            ".. we download faster than we upload because of economics, not physics. The economics are those of telephone and cable companies. Their business models assumed the value of the Net came from the content they deliver to us. The business models got written into a physical infrastructure that favors downloads over uploads.

            ... “Asymmetry is a belief system. The purveyors of connections looked at the Internet and saw TV, then acted according to what they saw.” "
          Your message has been successfully submitted and would be delivered to recipients shortly.