Cell Relay Archive

[mukul@trillium.com (Mukul Katiyar)]

Hi Fergal,

>
> I use private NSAP ATM addresses to connect to any remote ATM end stations
> on my private ATM LAN.
>
> I'm not to familar with public ATM addressing, that is, E164 addressing..
>
> If my application needed to communicate with an end station located in
> a public ATM network, what would it do...
>
> Would it encode the server's public E164 address in a NSAP E164 address ?
> and then would an interworking device at the edge of the private ATM network
> convert this NSAP address into a public E164 address...?? Or is there
> an alternative approach ??

 This is correct approach to address a host in Public ATM network from a
 device in the Private netwrok supporting only NSAP Address formats.

 Public networks may also support the Private ATM address structure
 ( depends on the network provider). In this case the ATM Address of the
 destination shall determine the type of private ATM address format to be used.



>
> So, What is the story with E164 addresses...
> How long are they... NSAP are **always** 20 bytes...
Yes, Private ATM Addresses are of the NSAP format and
are 20 octet long. Padding is used for any address that is shorter
than the given 20 octets, for padding details you can refer to
ATMF-UNI3.1 specs.

> What about E164 address lengths.. What is the max and what is the min ?
> I've read that they are fifteen digits... so would it be correct to
> say that they are **always 15 octets*** in length ?

They are 15 digits and encoded as 'BCD' enabling them to be encoded
in 8 octets with the last nibble of the 8th octet ecoded to 1111.
Also if say a number is less then 15 digits ( say a telephone number
which is lesseer then 15 digits ) then an E.164 address is composed
from it by padding the leading semi octets with 0000 to obtain
the maximum length of 15 digits.



>
> Where can I find a good reference when considering all of this ?

 most of the above information can be found in the E.164 recommendation
 of ITU-T and UNI3.1 recommendation of ATMF.

 Hope this helps

-mukul


>
> Cheers for any responses... (and please cc me in the response)
>
> Regards,
> Fergal.
> IONA Technologies.
> (NASDAQ:IONAY)