The Wizard of Oz
Q: How can operators deal with data usage that rises so much faster than revenue?
There are a number of strategies that people are adopting to try to make their networks more efficient. There are offload strategies, there are various network optimisation strategies and there are other ways, like increasing spectrum holdings. There are also moves you can make in terms of pricing the network for value—being able to offer differentiated quality of service and being able to charge users based on their consumption, which fortunately is something we’ve done since the beginning, both on fixed and mobile.
So the answer is basically a ruthless driving of efficiency while at the same time creating the business models that enable value to be extracted from the network in proportion to that which the user is getting.
There were a lot of “all you can eat” models in the industry, which is one of the reasons revenues weren’t keeping pace with the growth in capex demand—and I think you’ll start to see the industry correct it. Most carriers globally are now introducing capped plans. We’ve had those in place the whole time, but it’s still not enough. You’ve got to deploy new solutions like policy management and DPI.
Q: So do you believe that content providers and OTT players generating large amounts of traffic should compensate operators for its delivery?
I think my answer to that would be fairly predictable! I work for a carrier, and we clearly believe that this is how it should be. I think, though, that what’s going to happen is that we’re going to have to provide a way of getting better in our delivery of their services to customers. I think that’s the real key, here.
Q: Of those services, which are the ones that give your team the biggest headaches?
A large proportion of the data on our network comes from dongles and, in terms of traffic, the bigger the screen, the more traffic you drive. Interestingly enough, people with tablets don’t tend to use the 3G network as much, and that’s probably because there’s a different usage paradigm for tablets. The other thing that causes headaches are the services that drive signalling. So you get badly behaved apps that are constantly polling or making demands on the signalling network. Fortunately they’re also running people’s batteries flat, so people are learning to become more circumspect with those types of apps. I think that will probably get better with time as app developers learn not to do those sorts of things.
Q: What trends in capex and opex are you seeing as you build out the network to cope with all of this data traffic?
The capex is being driven by demand and there is no question that the massive demand on the network is driving capex at a much faster rate than we’d like. Opex on the other hand is being driven by the productivity gains of new technologies, so that’s tended to decline.
Cost management is a major issue because of the rising capex and the demand on traffic. In terms of the sales and marketing, everyone understands that users need and want data and you want to deliver it to them with the best possible experience. But obviously you have to do that in the most profitable way and it’s that that’s putting all operators under pressure.
Q: Speaking of productivity gains, how close is your network to the limits of spectral efficiency?
No network is ever really at the limit of spectral efficiency. Spectral efficiency continues to improve with advances in tenant techniques and in multipleinput and multiple-output (MIMO) and multi-user antennae technologies. Do we think based on today’s technology there’s a huge amount of spectral efficiency gain? There’s some when you go to LTE, which is one of the reasons we’ve deployed LTE for capacity enhancement. But put it like this, we’re using spectrum as efficiently as anyone could with the current state of technology.
Q: What can you tell us about the status of your LTE deployment?
We’ve deployed an LTE network at 1800MHz in re-farmed GSM spectrum and we’ve deployed it as an alternative to building out extra 3G capacity. We’ve started using LTE as a way of increasing capacity— it covers around 45 per cent of the population now. It’s not a full coverage network, it’s deployed mainly around the major cities in a radius from the centre—and in airports and places where we have significant capacity issues.
Q: How have spectrum licence requirements changed with LTE?
Our spectrum regulation environment is probably one of the most sophisticated in the world; in fact we’ve tended to lead the world in a lot of things. For example, we were one of the first countries to use spectrum licences rather than apparatus licences, which enables you to buy spectrum and deploy using any technology that sits within the power profile and within certain regulatory requirements. You couldn’t buy broadcast spectrum and then deploy a mobile network in it because the way in which broadcast and mobile is licensed is different. What you could do is take your GSM spectrum and use it for 3G; there’s nothing in our spectrum environment to stop us doing that. And over the past couple of years, as the government has moved towards auctioning off the new spectrum bands, the digital dividend band at 700MHz and the LTE band at 2.6GHz, they’ve also been renewing the spectrum licences which are reaching the end of their term. As they’ve done this they have been introducing more sophisticated rules around the use of spectrum. One of the things which we favoured was a “use it or lose it” rule, so they are getting tighter on operators’ efficiency of spectrum management.
Q: With this in mind, how much longer do you expect to be using your GSM network?
That’s a tricky one because in-bound roaming will probably require it to be in use for quite a long time. The other thing is there are telematics devices that use GSM. M2M is a different sort of market because devices stay in use for a very long time, so I don’t see the GSM network going away any time soon. I can see a time in the future where we may be able to free up 5MHz to use on 3G or 4G. But the fact is that the GSM network is not a separate network per se, it’s built into the 3G network.
Q: With regards to vendors, is there a big difference in terms of product quality?
No—vendors have different strengths in different fields. You’ll have one vendor that is very strong in the mobile area but not as strong in the broadband area and another one where the strengths are reversed. Some of our vendors have stumbled on certain deployments but been very good in others. What is important, though, is that we tend to work with big vendors that are very intimately involved with us in the sense that we have strong working relationships. So I wouldn’t say there’s a huge difference between them.
When selecting a vendor, our criteria is more around partnerships—we’re trying to work in a more productive way with vendors under more of a partnership arrangement rather than purely a transactional arrangement. Aside from that, factors like their technology roadmaps, their position in the industry, those make a big difference. Our opinions about the future performance of suppliers certainly has an influence [on our decisions] because we are looking for the strongest roadmap. So if you see suppliers struggling to get the finance to keep their roadmap up to date and stay leading edge, that would definitely influence your selection criteria.
Q: Is there still a place for smaller, specialist vendors, or do the large players have everything covered?
There has to be a combination of both. The problem that niche companies have, and I’ve worked with many over my years at Telstra, is the cost of sales to a large telco like us. We have very stringent requirements in terms of supportability and maintenance and all those sorts of things. But in new markets, we do tend to go with new companies. Some of the media markets we’ve gone into in the past, such as our hybrid TV service, the set top box comes from a small firm. We weren’t the first large telco to use them, they have a customer in their home market, but it does tend to be the new players that introduce new technology into a metric. Q: If you were investing in a technology startup, where would you look to innovate?
I’d probably try to do something in the space of using network intelligence to optimise the delivery of service to customers. There are companies that are doing this today but I think that anything that will enable you to better utilise your network resources by using the state of knowledge of the network itself, would be a fruitful area. Something that will be able to take network information and either give you better tailored customer service or give you better efficiency in terms of traffic delivery.