On 10/01/2018 4:12 PM, David wrote:
> On Wednesday, 10 January 2018 2:34:20 PM AEDT Paul Brooks wrote:
>
>> All of these [voice transport protocols] are part of the PSTN, and the 
>> traditional telephone number addressing is the Universal Voice Glue that 
>> makes it all work together relatively seamlessly, and lets someone on a POTS 
>> line call someone on IP line without having to be aware of what technology 
>> the receiver is using at that moment. Not quite irrelevant!.
> However IP-based voice connections will be in a huge majority when the NBN 
> finishes its rollout, probably enough to warrant a strategic rethink.  Call 
> routing based on POTS number, as currently done by the carriers, logically 
> duplicates a function which could, in principle, be done by the IP (i.e. NBN) 
> network.  Surely it should be possible to integrate the IP and remaining POTS 
> networks so the whole system is more efficient and way less cumbersome.
I think what you're looking for is RFC 6116 ENUM - a DNS lookup of a telephone 
number
to a URI such as a SIP address, and RFC 5067, RFC 5526 "Infrastructure ENUM", 
and RFC
3824 "Using E.164 number with SIP".

Check https://en.wikipedia.org/wiki/Telephone_number_mapping for an overview.

However, having a lot of IP-based access lines is not the same as having a 
world, or
even a country, where everybody can be interconnected end-to-end with a VoIP 
call.

Considerations like lawful interception, metadata retention etc require the 
call to
pass through the RSPs on the way through, it can't just take a shorter routing 
path
direct IP-to-IP or neighbour-to-neighbour like nirvana of IP telephony might 
imagine,
as it might occur within a single campus like a corporate office, so there's 
little
real routing efficiency to be gained.

Sure routing by POTS number duplicates a function that IP networks could do, 
for calls
to another IP line - but the telephone service transcends IP and exists above 
IP, in
an application layer, and needs to work across all the other technologies as 
well - so
it should use an addressing and routing system appropriate to its layer, not be
limited by the shoehorning it into a lower layer. ENUM is the translation 
method from
one to the other.  The reality is that a large number of calls to and from an
IP-enable voice line won't have another IP-enabled voice line at the other end,
they'll hit a gateway in the middle. Incoming calls from international 
locations,
corporate offices and call centres (generally on ISDN-PRI-based PABXs ), mobile
networks and conventional POTS lines will all still need to use a E.164 
telephone
number to reach the consumer.

From the consumer making an outbound call, its still actually easier and 
quicker to
punch in a telephone number than to tap out a long SIP URI, and devices need a 
much
smaller keypad too, and for those using a directory where they select a name or 
a face
(like a mobile handset contacts directory) it doesn't really matter if the 
underlying
entry for the picture is a number or a SIP URI. So from a usability 
perspective, the
conventional telephone number is still actually easier to use.

(Also, re the NBN - The NBN is an Ethernet Layer 2 network, not an IP network, 
and
consists of 121 disconnected islands rather than a national network, so 
national VoIP
interconnection can't and shouldn't be done within the NBN anyway!)

cheers,
    Paul.


 




>
>
>>> I see, I'd assumed the mandatory requirement to publish a PLNR ("Ported 
>>> Local Number Register") file was intended to allow all carriers to route 
>>> calls directly to the carrier currently holding a ported number without 
>>> going through the donor carrier.  But the whole idea might be suffering 
>>> scability problems now.
>> That is the idea - but also, even interconnect arrangements are bilateral as 
>> well.  'Your' carrier may not even have a bilateral network interconnect 
>> directly with the final destination network hosting the number, and may have 
>> to route the call through a third network who will provide the transit 
>> connectivity (who might or might not be the original donor carrier) - who 
>> will do a second lookup of the PLNR in the process to work out which 
>> direction to forward the call to.
> Out of interest I downloaded the full "EnhancedFullDownload.csv" file (81.4 
> Mbytes!) from
> https://www.thenumberingsystem.com.au/#/number-register/search  This shows 
> the allocated and current holders of the entire POTS numbering range, but at 
> a quick look I couldn't see any individual numbers, just ranges.
>
> It's all something of a mystery, I suppose there's probably a degree of 
> ad-hocery going on...
>
> David L.
>



_______________________________________________
Link mailing list
[email protected]
http://mailman.anu.edu.au/mailman/listinfo/link

Reply via email to