Advertisement
If you have a new account but are having problems posting or verifying your account, please email us on hello@boards.ie for help. Thanks :)
Hello all! Please ensure that you are posting a new thread or question in the appropriate forum. The Feedback forum is overwhelmed with questions that are having to be moved elsewhere. If you need help to verify your account contact hello@boards.ie

Contention Ratios - Mathematical Analysis

  • 18-01-2004 4:06pm
    #1
    Reyman
    Registered Users, Registered Users 2 Posts: 1,297 ✭✭✭


    Has anyone any detailed knowledge of how to tie the concept of contention ratios to the true QoS experienced on a line? I'm thinking beyond for example - CR 20:1 implies a minimum speed of 512Kbit/s /20 = 25.6Kbit/s etc.
    But to the example of voice where a Poisson distribution is assumed and Erlangian calculations allow easy evaluation of the probability of completing a call eg 95% or 99% etc.
    In the case of contention ratios is there any mathematical way of arriving at a similar confidence level for a particular required speed of service? I'd like to have some solid information or 'white paper' references in dealing with my good friends eircom!


Comments

  • #2
    longword
    Closed Accounts Posts: 741 ✭✭✭longword


    I'm not sure it would be very useful. How badly the contention bites will depend on how complete the contention is - often there won't actually be another 23 bodies signed up to share with you (Eircom's Solo and better packages are 24:1, RADSL is 48:1, and ESAT's own service (not IOL) is 20:1).

    It will depend on whether or not the company has tried to discourage excessive usage, the nature of their customers, and whether or not by chance you've happened to wind up sharing with two or three hogs. It will depend on whether there's been a new Windows service pack, or a new Linux distribution released this week.

    The variation could be huge and unpredictable.


  • #3
    zoro
    Registered Users, Registered Users 2 Posts: 2,243 ✭✭✭zoro


    would that mathematical formula also take into account the fact that between 9-5 there's generally no one home to use the service? im thinking along the lines of a family's usage of the net, wihtout a very technically minded child / parent in the house

    they'll generally only use it to check the mail, pay the bills, banking online, maybe chat to some friends / relatives abroad or something you know?

    you could say that between 9-5 there's hardly a high chance of more than 3 or 4 people contending on any one line at once.

    then again you could have 16 warez whores on 24/7 on your connection you know?


  • #4
    Reyman
    Registered Users, Registered Users 2 Posts: 1,297 ✭✭✭Reyman


    Thanks all ! But what i'm looking for is some kind of way to come up with confidence levels for particular access speeds eg. 128Kbit/s ...95% certainty. This is quite well documented for voice
    and you never get congestion

    The planners of this BB network must have some means of doing this type of calculation to protect them from the wrath of their customers.

    On second thoughts maybe that's not a big concern!


Advertisement