Why Telstra is wrong on VoIP (by Simon Hackett)

Written by Renai LeMay on Wednesday, May 4, posted on Delimiter

In this opinion piece, Internode managing director responds to comments by Telstra executives earlier on this week about Voice over Internet Protocol-based telephony.

opinion Reading this article, there is just so much wrong with the reported comments from Telstra here about VoIP. Its got technical inaccuracies and factual errors of significant sorts all through it.

I’ll work through a representative sample of what I mean …

To ensure the service provides an acceptable level of quality, Telstra has pledged to spend part of a $600 million package on upgrading equipment in its telephone exchanges with Broadsoft hardware to support quality of service (QoS) techniques to prioritise voice traffic.

Broadsoft make the VoIP soft-switch that handles call switching decisions for VoIP traffic. They don’t make network hardware. So the wrong vendor has been named here in this context. Its like saying that you’re getting a tyre manufacturer to improve the way a plane’s wings work. This speaks to the general lack of accuracy in the statements made.

Next, and more tellingly, when Telstra speak of upgrades needed to make VoIP work properly, its important to understand that this means that Telstra have (clearly) under-invested to date in their ADSL network and its management of QoS in both the backhaul and DSLAM components of their network — but that existing VoIP providers with their own DSLAM networks (such as Internode and iINet) designed in the appropriate QoS right up front.

So the key point to understand is that when Telstra say that VoIP isn’t yet reliable, they are making a statement about the quality of the Telstra network only, not about VoIP in general.

They are saying their network isn’t up to scratch (and apparently requires $600 million to fix it), not that VoIP in general isn’t up to scratch. And clearly the many operators of VoIP services providing them to hundreds of thousands of Australians don’t have the problems with it (nor do their customers) that Telstra feels they should be having.

It’s also surprising to see that $600 million figure in the context of this being a network Telstra intend to shut down in the National Broadband Network era. What are they really spending that money on? Is it really network upgrades, or is that just the size of their marketing/rebate slush fund to try to draw customers back from competitors with subsidised deals? Is this sort of margin erosion the best course of action for Telstra to use to best benefit its shareholders?

Hence it seems that the correct re-interpretation of “we don’t think the quality and reliability is there. We could bring it to the market tomorrow, but we don’t want to” is really “our network isn’t up to scratch as yet, unlike our competitors”.

Next comment to respond to:

Telstra’s small business chief Deena Shiff said Telstra’s ‘voice over broadband’ solution was qualitatively different from iiNet’s consumer-grade VoIP, as Telstra was investing to build quality of service into its exchanges – whereas iiNet’s solution relied only on QoS embedded in its routers on users’ premises.

This is the same mis-statement. iiNet’s network (and Internode’s network) were built with designed-in QoS and with appropriate investments in backhaul capacity so that the quality of VoIP services on those networks is just great. Routinely higher quality than, say, a mobile phone network. In other words, the investments Telstra says it’s making to bring in ‘Quality of Service’ are already done in their competitors’ networks. They’re not doing something others haven’t. They’re merely catching up from a current stance of being far behind in this important area.

“They don’t give you an end to end quality experience,” she said. Shiff emphasised that Telstra’s solution was “not VoIP”, but instead described it as “digital voice”, stressing that the high definition of the audio set the Telstra solution apart.

There are two distinct points to respond to in the quote noted above.

1. Of course the Telstra solution is VoIP. Telstra have named Broadsoft — whose product, Broadworks, is a VoIP soft-switch. This is the very same (high end, high quality) VoIP switching software product that is in use in the national Internode and iiNet VoIP service networks today, and that has been in use in both of those networks since circa 2005. Broadworks don’t make some magic pudding called ‘Digital Voice’ that is not VoIP. They’re one and the same thing.

This sort of sleight of hand is disappointing to see, as its just not an honest representation of a situation where Telstra are catching up to the rest of the market (or rather, are promising that they’ll do so). They’re launching a service using this ‘digital voice’ in a month. Are they really spending $600 million on upgrades in just 30 days, or again, is this figure really just the size of their marketing budget?

2. “High definition audio” is a feature of the current generation “Fritz!Box 7390″ and ‘Fritz!Box 7270″ Home/SME ADSL2+ routers that are sold by Internode today.

HD Audio is a standard capability of VoIP services using sufficiently good VoIP hardware (and the Fritz!Box products are simply at the top of the heap in this regard). The quality delivered by a Fritz!Box using its DECT cordless handsets, when calling another HD Audio endpoint (another Fritz!Box or otherwise) is so good that it’s really quite spooky — far and away better in quality than a traditional PSTN phone call. And of course dramatically lower cost (for instance, unlimited NodePhone VoIP to NodePhone VoIP calls on the Internode service are free of per-call charges entirely).

HD Audio uses a standard VoIP CODEC (G.722). This is not a Telstra innovation in any sense. It is, again, Telstra catching up, late and last, with capabilities that its competitors in this space have been offering to the market already.

Finally, and a key point here — while Telstra are framing VoIP as being somehow inadequate for consumer requirements, they are deeply in negotiation with NBN Co to turn off their copper network and move all their voice endpoints nationally over to the NBN. Guess what — the voice ports on the NBN customer termination boxes are actually … VoIP hardware. On the NBN, your “PSTN” service will be turned into VoIP right in the NBN Co fibre termination box and the call will be carried via standard VoIP/SIP protocols — exactly as all VoIP providers already do today.

There is a deep sense in which the statements made about VoIP by Telstra today are simply trying to deflect its own status as the last adopter of this technology by claiming that it is somehow not going to be good enough until Telstra ‘invents’ it via some mysterious magical property imbued upon it by calling it ‘digital voice’ instead of VoIP.

It’s like IP Multicast — a technology that Internode and others use to deliver highly efficient linear TV channels using our FetchTV service. We can’t sell that to customers of ours that we reach via Telstra Wholesale ADSL2+ services, because Telstra haven’t yet seen fit to ‘invent’ IP Multicast. So in their world, it doesn’t exist.

Sorry, but it does. And so does High Definition, high quality VoIP services. Sure, you have to use appropriate hardware at the customer end point (like a Fritz!Box) and you need your network to be built properly.

But everyone in the industry ‘cept Telstra already does these things today.

Image credit: Internode