Opinion: Time to re-think telephone numbers

ACMA needs to widen its scope, writes David Havyatt.

As the ACMA commences a review of Australia's telephone numbering plan, David Havyatt wonders whether the scope of this review is broad enough.

The ACMA has commenced a wide ranging inquiry into Australia's telephone numbering scheme, starting with the release of one discussion paper and a promise of three more to come. 

The ACMA identifies a number of motives in undertaking the review. One is the observation that a telephone number used for a VoIP service is only an address and not an identifier of a specific physical connection.

Another is the recent publicity of the fact that "freephone" and "local rate" numbers (1800 and 1300) are incorrect designations descriptions of these services, as they cost a lot more to call from a mobile.

The challenge from a policy standpoint is that these "problems" extend to other parts of the regulatory framework - to areas outside the responsibility of the ACMA.

The history of numbers

We take telephone numbers for granted these days, as if it is natural that you need to dial a number. But increasingly we use internet services that use names (e.g. URLs and e-mail addresses) or rely upon directories stored locally to use a name to make a connection.

In some ways it's a return to the earliest methods of telephony. You asked the local switchboard operator for the person by name. It was only the advent of automatic switching that introduced the use of numbers for different exchanges. The advent of automated long distance calling required the development of a numbering plan covering the whole country, which was one component of the Community Telephone Plan of 1960. 

But by the early 1980s, Telecom broke down the original hierarchical numbering structure. Number length varied in a piecemeal approach, and numbers in 1000 number blocks started appearing on different exchanges with the same prefix.

Number portability broke this down further with the consequence that numbers could now appear on someone else's network. 

This is one factor that leads the ACMA to start speculating that numbers are losing their geographic significance. The second is that the effect of competition has largely collapsed the long distance charging bands, and that many plans, especially VoIP services, make no local versus long distance distinction. 

But the integrity of the geographic information is retained by the numbering plan itself.  Numbers are limited to charge zones - and charge zones determine what is, and what isn't, a local call. 

In Australia, the right to local calls is protected in legislation under a convoluted definition that ultimately rests on decisions made 50 years ago. In essence, a call is a local call if it is to another person in the same charge zone or in an adjacent charge zone.

The trouble for the numbering plan is that this right applies not only to those with a VoIP service but to people ringing someone with a VoIP service. So it is not possible to move away from "geographic" numbers for voice services without legislative change.

Early attempts

The ACMA's first attempt to address the issue of numbers and internet supported services was the ENUM project, which is now consigned to merely a footnote in the discussion paper. 

This was an IETF attempt to create a map from ITU-E.164 (telephone) numbers to various IP addresses using the DNS [Directory Name Server]. In effect a URL was created for each number so that the DNS could retain information about that number, most notably its IP address for a VoIP service.

The ACMA has also attempted to create a new number range for "nomadic" numbers - for VoIP numbers of no fixed location. However, numbers are only one part of the regulatory framework for creating services. They also need to be incorporated in the interconnect standards. An important part of those standards is not just information to be carried in signalling, but at which Point of Interconnect (POI) the call is handed over. 

As a simple example, if I am in Sydney and I call a customer of another network who is in Perth using their geographic number, my provider carries the call to the POI in Perth. But if I'm calling the person (in Perth) on a freephone or local rate number, my provider hands the call over in Sydney (my provider doesn't know they are in Perth).

The second difference is in the way the money flows. In the first case, my provider pays the second provider for terminating the call, while in the second case my provider is paid for originating the call. 

Without working out both interconnect technical and commercial regimes, a new number range is meaningless.

This overlap between numbering and interconnect lies at the heart of the mobile and freephone problem. There used to be the appropriate "declared service" - but the ACCC revoked the declaration because it was unused.

While the ACMA is dedicating a great deal of effort to considerations of telephone numbering, it only mentions in passing the issue of IPv4 numbering expiry.

It is their position that the IPv4 problem will be resolved by industry when the matter needs to be, a position that possibly over-relies on the market. Those with market power through control of the declining pool of IPv4 addresses have incentives to maintain that control, and to frustrate the utility of end-to-end applications. But that is a whole topic in itself. 

This reveals a weakness of the ACMA's conception of the policy reasons for numbering administration being a government function. The discussion paper identifies competition as one of the policy objectives, but limits this to the role of number portability in facilitating choice of provider. 

However, simply guaranteeing access to numbers is an important part of facilitating competition. The introduction of competition in long distance in North America in the 1980s meant that users had to dial some incredibly convoluted numbers to choose a carrier and then dial another party - as numbering was controlled by incumbents. The competition objective was the primary reason why Telecom Australia lost numbering responsibility to AUSTEL in 1989.

The initiative of the ACMA in launching this latest inquiry into numbering is to be lauded. However, numbering is too intricately linked to other regulatory elements for the ACMA to deal with the issues alone. 

The Communications Minister, for example,  has previously committed during the election to a Convergent Communications Review and last month told the National Commercial Radio Conference that the review will consider how the regulatory framework applies to communications services in a converged environment.  The review will also seek to identify appropriate licensing regulations, regulatory obligations and consumer protection arrangements across a range of media platforms, including radio."

It is unclear at this stage whether the scope is limited to "media platforms" or addresses all regulatory aspects.

It would seem that the future of all numbering, naming and addressing schemes should form part of the review.