Kathy Gibson reports from SATNAC 2014 – The telecommunications industry – both in Africa and around the world – is at a turning point.
That’s the word from Jean-Pierre Lartigue, head of corporate strategy at Alcatel-Lucent, speaking at today’s SATNAC 2014 conference.
He says that, while the world has changed dramatically over the last 10 years, it will change even more over the next 10.
“As users, we have been changed by technology,” he says. “Our life is now a digital life and we are immersed in technology – an extension of our life is in the digital world.”
Connectivity doesn’t just benefit users though. Lartigue points out that connected countries can take part in the global economy.
“We see a cycle of investment, starting with broadband connections. A supply strategy is the place to start and then build upon that.”
Thereafter users demand more so users connect globally so there is an ease of use and global digital reach; users change their usage and get social so you get a shift from local to global competition. When users digitalise and socialise everything there is a shift from the old to the new economy and a growth in jobs skills and GDP.
With massive growth anticipated over the next three to five years Lartigue says the solution of today cannot be the solution of tomorrow. We need to look at new technologies to drive growth.
“I think the network has to change to fit the user,” he says. “In the past, we put down infrastructure and users came to use it as it is – users had to adapt to the network.
“The reality today is that users want versatility and flexibility, the network has to adapt to the user. It has to become a transparent skin that follows us and adapts to what we want personally. This will help the network to cope with the vagaries and scale of what is needed.”
For this to happen he says the network needs to change on five dimensions:
* The network will be faster – it has to be economical and allow connectivity, but it has to be faster, so it has to break the limits of physics. “They are there to be broken and our scientists are breaking them every day
* The network will be cloud-centric – the physical infrastructure falls away, with a fusion of dynamic cloud and programmable network infrastructure working together. Multi-service, multi-tenant capabilities are coming to market that challenge the idea of networking infrastructure. There is a rapid transformation of networks into a cloud infrastructure. This makes the network more fluid.
* The network needs to get closer to the end user. We are seeing fibre getting closer to the user and now content delivery and cloud functions are also getting closer to the users. This means the cloud is becoming more distributed, serving small pockets of the population. What we seeing for the future is a deep transformation of the topology of the network, with mini-clouds close to users. This requires a change of cities and neighbourhoods.
* The network is user-centric and needs to become user-aware – reaching out to us to understand how we use it so it can priorities and tailor services. This will drive network-enabled machines and network-enabled video and cloud, also to scale on traffic and select only the services needed by that specific user.
* The network has to be secure. All the data that the network and service brings needs to be security. “Security is a big topic for us, and a big challenges for the whole industry,” says Lartigue. “Collaboration between enterprises, consumers and governments is critical.”
So clearly there is a turning point now, and a change in the industry, there are dramatic breakthroughs that we need to think about, says Lartigue. The new network needs to be open, with cross-industry collaboration; versatile with distributed cloud networks; and rapid with service creation and premiums.
When this happens, Lartigue says, the network will realise its true value.
Cassie Bezuidenhout, consulting system engineer at Cisco, agrees that, as we move to a world where billions of things are connected, the network will have to undergo some massive changes.
He points out that, by 2018, global IP traffic will reach about 1,6 zettabytes. By 2018 there will be close to 4-billion people on the Internet, with more than 21-billion devices connected as well.
This means we need faster broadband speeds, and the need for more speed just gets higher all the time, Bezuidenhout says.
The challenges for operators, he says, is to reduce network costs while simplifying and streamlining operation models, Bringing out new services quickly is important, but existing service level agreements have to be maintained or even exceeded at the same time.
But this is not just a problem for the service providers, Bezuidenhout says, but for the vendors who need to work at reducing the cost per bit.
Today’s networks have come a long way in the last 10 years, he says, but there are still big challenges that they are unable to overcome.
All of the available technologies and breakthroughs need to be combined on a single platform, Bezuidenhout says.
“But the real problem lies with power and cooling. It currently has a high cost, and the acoustics mean we will almost have to ship a fridge with every chassis.”
Vendors need to be smart to come up solutions to cooling, he adds. Integrating components, using optical interconnects, using optical switches, or providing front to rear air cooling could all provide savings.
But industry needs to come up with more, new ideas Bezuidenhout says.
Another challenge is the fact that dynamic services require dynamic networks that need to be simplified, and solutions are needed in this regard.
Some ideas for simplification include leveraging software-defined networks (SDN) or leveraging hybrid SDN models. “But this needs to be vendor agnostic,” Bezuidenhout stresses.
“It’s an exciting time from a vendor point of view, with 50% of the world population currently connected and more devices set to be connected – more than half of which is mobile,” he says.
“We need to figure out how to continue to scale. The network is the enabler, and the application needs to drive the network. To do this, do we continue to build networks the way we used to, or do we change? And what is the solution?”