Three TDs have been uploaded
okubo at giti.or.jp
Thu Sep 3 00:17:51 EDT 1998
It's a shame I wasn't on the call, I might find out what's going on :-(
Thanks to those who are exposing it on the list. :-)
In my mind, address resolution (or number translation) is a multistage
In the context of the caller, a destination address may be an "internal"
form like an abbreviated dialing number (*21), a service prefixed number
(01161294332374), or a nickname ("doug"), that has to be translated into an
"external" form. The result of this stage is an external address.
Once in the external form that identifies the callee, the number is subject
to resolution in the context of the callee. I think that 800 numbers are
like this. Since geographical information is probably not known about the
source or the destination, much less path metrics between them, factoring
that into the resolution may be a problem. The result of this stage is an
If the response is a prioritised sequence of endpoints, then the source can
select from the endpoints within each priority based on characteristics.
For example 1-800-555-1212 => (1-415-555-1212, 1-212-555-1212) and I can
select my area code if I know it.
The third step involves endpoint location, sometimes called gateway
location. This involves taking the effective address and locating a
physical (transport) address that can handle it. This could also be a
prioritised sequence of endpoints (gateways), for each effective address.
The source can then select from the endpoints within each priority based on
characteristics. I might look in my IP routing tables, ping them or use RSVP.
I don't think that "geographical" proximity is the *real* determining
factor here. It doesn't matter if the source is sitting on top of the
destination if the link is a 300 baud serial line.
I think that the problem being canvased here is one that comes from the
PSTN domain where the 800 charges are distance dependent. I don't think
that it translates directly to the packet network domain, where charges are
often distance independent.
What's more likely to be interesting, is the moving call centre where the
effective address follows the sun from London to San Francisco to Sydney to
maximise daytime staff utilisation. Here your 1-800 number will translate
to 44-171, 1-415, and 61-2 depending on the time of day.
At 18:01 02/09/98 +0100, Chris wrote:
>Good point, Pete! It's a shame you weren't on the call (there was a
>certain amount of guesswork as to how 800 numbers work)!
>There are two ways of handling this that I can see immediately. Although
>philosophically different, in practice they will normally amount to the
>1. The "owner" of a domain needs to be told where a query originates,
>given that the calling location may affect the answer it gives or:
>2. A border element handling queries for 800-type numbers needs to know
>what to do with it depending on the origin of the query.
>Dr Chris Purvis - Senior Development Engineer, WAVE CC Software
>Madge Networks Ltd, Wexham Springs, Framewood Road, Wexham, Berks.
>Phone:+44 1753 661359 email: cpurvis at madge.com
>It may not be possible to handle 800 numbers in the same way as other
>numbers as the destination may be based on the geographical location of
>the caller. For example, ISPs may be accessible by something like an
>800 number which may be fixed for a whole country, but the POP you end
>up connected to may vary depending on where you called from.
>I'm not sure what the implications of this are, but it is something to
>E-Mail: pete.cordell at bt-sys.bt.co.uk
>Tel: +44 1473 646436
>Fax: +44 1473 645499
>>From: Chris Purvis WVdevmt-WS[SMTP:cpurvis at MADGE.COM]
>>Sent: 02 September 1998 13:09
>>To: ITU-SG16 at MAILBAG.INTEL.COM
>>Subject: Re: H.323 Addresses
>>>We are discussing about H.323 addresing schemes in our bi-weekly H.323
>>>inter-GK conf calls.
>>>The H.323 addresses that are being considered are 1. E.164, 2. E-mail,
>>>URL, and 4. TCP/UDP/RTP port addresses (and 5. aliases as well I
>>>Although a hierarchical notion of addressing scehmes have been
>>>have also recognized that some other variations of addressing schemes
>>>For example, people change their physical locations, but they might
>>>keep the same E.164 addresses. So, a translation is needed. Therefore,
>>>very physical relationship of E.164 addressing scheme (e.g., knowing
>>>and NXXs, one can find the distance) has been broken.
>>I was under the impression that finally (after several weeks of
>>discussion) we'd reached closure on this in last Thursday's call. The
>>solution that I understood had been agreed came in two parts:
>>1. Mobile IP solves most of the problem for us - E.164 number resolves
>>an IP address which may be anywhere in the world at any given moment in
>>2. If a user wants to move (and take their number with them) to being
>>an administrative domain with an otherwise reasonably contiguous E.164
>>address space, this is easily handled as follows. The domain that
>>"should" (in the hierarchical sense) "own" the number that the user is
>>keeping holds a record of where to contact the current holder of that
>>number (which will typically be an entity to which an LRQ message can
>>>Another example can be 800 numbers where the translation is also
>>It seemed to be agreed in the call that 800 numbers are not a special
>>case, but merely an example of number portability.
>>>In addition, from mobility point of view (device or person) wired (and
>>>wireless?) environement may also be looked into.
>>Does this mean anything at all?
>>>As we go froward, we may consider many of those aspects as pointed
>>>provide solutions for the H.323 addressing schemes.
>>I was under the impression that we'd reached closure on most of these
>>Dr Chris Purvis - Senior Development Engineer, WAVE CC Software
>>Madge Networks Ltd, Wexham Springs, Framewood Road, Wexham, Berks.
>>Phone:+44 1753 661359 email: cpurvis at madge.com
More information about the sg16-avd