Chatting with Huawei this week we came across an interesting set of ideas from Richard Brennan, Vice Director of Industry Standards at Huawei, who was talking about the future for broadband and its charging mechanisms.
While he was at pains to point out that this was only his opinion and not a strategy that Huawei pushed on its operators, he put forward the simple idea that if net neutrality and Over The Top services were to both happen, then instead of stopping them, operators would need to come up with a plan for making money out of them.
While there is a long way to go on this, and this is not a global proposition, because in one country net neutrality may be frowned on, while in another it may be embraced, Brennan’s ideas have come from talking to countless operators and distilled from their ‘worst case’ strategies.
In essence instead of broadband operators trying to get extra money from companies like Google for better consumer access to their web properties, the proposition needs to be broken into different delivery bands at the customer end.
‘It will be dressed up in marketing, and could come out as Silver, Gold and Platinum services, but in the end there are so many services that will be going into a home that the operator cannot provide them all and has to work with other providers. What they can do is provide a different QoS policy from household to household, service to service, and consumers can choose whether or not to pay for that.’
Currently most ISP thinking is focused on getting Sony to pay for its game bits, Google to pay for its search bits and Netflix to simply not get blocked out of existence because it takes so many bits. But another way of looking at this problem is to offer a customer defined QoS or Quality of Experience (QoE) promise either to the entire household, or better still to the particular services. At its most granular it could be to some services, some of the time.
That way an ISP could work in co-operation with Google or Sony or Xbox or Netflix. ‘Certainly madam you can have the new Netflix service ‘ you’ll get the best reception if you upgrade to the Gold AT&T service, would you like me to arrange that for you?’ or some such, with a kick back for the service upgrade.
Remember one of these players could easily be the backhaul for femtocells owned by Cellcos, and they are a new element in this QoS policy management equation that may be as contentious as pirate video, not because it eats so much of the bandwidth (although it may) but because it could be a direct rival to the ISP. Telefonica would be unhappy if in Spain on its fixed broadband lines it was backhauling Vodafone cellular traffic.
‘It could be even more granular than simple gold or platinum services,’ suggests Brennan. ‘If you decide to get your Disney Over The Top, direct from Disney, that should not imply lousy reception, so you should be able to receive an experience which offers equally good viewing as on broadcast TV. And you may want different reception for different programs. For instance the category of Live Sport might have a different QoE to any other TV programming, and operators need to solve the purely marketing problems of how to offer this in sensible packages for the consumer.’
Extrapolating from Brennan’s ideas we can see that the world suddenly becomes a less fractious place and partnerships become the name of the game. Partnerships between sports franchises, broadcast and cable channels, as well as with pay TV operators, all cross selling one another’s products.
If TV Everywhere becomes widely accepted, then the programs you have already paid a pay TV company for, can be delivered over the internet or the pay TV network. In fact in cable this is often a different protocol on the same network. Working out what bandwidth policy to apply to those programs should be equally beneficial to the process of offering better pay TV.
The other discussion we had with Brennan related to next generation DSL. Huawei is one of three companies (the other are Nokia Siemens Networks and Alcatel Lucent) which have demonstrated lab versions of DSL at the 700 mbps level or higher. Actually the Huawei version highlights more about MIMO than Phantom parallelism and calls itself Super Mimo and uses the 4 separate cables in two twisted pairs of telco wires.
‘It’s premature to think in terms of putting these techniques into existing DSLAMs,’ he said, ‘we have to lump phantom and MIMO approaches with all the other potential improvements and come up with a standard, before operators are going to want to install what are laboratory speed tests.’
This is a process we feel will take at least 18 months to two years and recent suggestions from NSN that it would begin putting this capability into its DSLAMs within a year are perhaps fanciful. The timeframe of 2013 or beyond is a bitter prospect for Western Telcos, especially in the US where they are up against Cable operators who continue to push 150 Mbps broadband as a differentiator.
‘I don’t know about NSN, and of course an operator could choose to put proprietary DSL technology into its network, but they might have to pay a considerable premium in order to do that.’
This area is something of a race and an expensive one, with quite a lot to balance in the equation. A telco has already invested in successive DSL technologies, but what comes with that, in order to bring the DSLAM closer to the home, has been a lengthening of fiber.
AT&T with its project Lightspeed spent $billions getting 26 million US homes closer enough to its fiber that it could offer them its U-Verse IPTV.
‘If it turns out that next generation DSL means an even shorter loop length, then that’s what has to happen,’ says Brennan.
What that means is that where an operator can get back their money from such an investment, and for this it will need a clear business model at a long term, sustained pricing level, it can afford to put yet more fiber closer to homes, slowly shortening loop lengths through successive generations of technology. The final step, of course, is to put fiber into every home.
‘We may be going with these new MIMO/Phantom approaches down from working over 5,000 foot loops, to around 500 meters (roughly 1,500 feet),’ said Brennan.
Interestingly in Huawei’s native China this would probably make no difference. Most of rural China cannot hope to get down to loop lengths that make ADSL2+ effective (5,000 feet) so they would not be able to benefit from any newer technology. Some cannot even use long reach DSL technologies because they are so far out.
‘We don’t have very much infrastructure at all outside the Cities,’ Brennan affirmed.
But in those Cities they are built in the modern way with multiple metropolitan multi-gigabit Ethernet rings, and everything branching from there. They have been designed with modern DSL connections in mind.
So as we have pointed out in the past, flexible wireless architectures like WiMAX are the best hope for remote rural broadband, while even more homes will be left out of the next generation of DSL unless companies like AT&T yet again undergo more fiber extension programs like Project Lightspeed to reduce the loop once again.
Failure to do that will lead to a bigger and bigger digital divide between the broadband capabilities of cities compared to rural locations.
But failing to build out fiber will yet again mean the continuing losses of broadband customers from telcos to cable companies, so the choice is between spending $billions or losing $billions in revenues. Tough Choice.