Shared small cells, and exploiting the convergence of wired and wireless architectures, could reduce the cost of deploying high capacity access networks, according to Dr Tod Sizer, Head of Wireless Research at Bell Labs and the architect of Alcatel-Lucent’s lightRadio innovation.
Talking to The Mobile Network, Sizer outlined two key trends that the networks research team at Bell Labs is exploring. The first is driven by the fact that, according to Sizer, wireless and wireline networks are becoming more and more alike.
As Sizer put it, wireline networks are now more often than not connected over the last few metres by WiFi. At the same time, cellular networks are densifying, and require a lot of wired backhaul. So, in Sizer’s words, we are seeing “two types of networks which are now moving to the same sort of overall architecture: that of a highly sophisticated wireline network serving a short distance wireless network.”
In fact, Bell Labs has merged its wireline and wireless research teams, Sizer said, as a result of this convergence.
“We’re looking at how we might be able to exploit that [convergence of architecture] in order to seamlessly hand off between what has traditionally been in the cellular space and what has traditionally been in the wireline space.
“The fact is that on WiFi, performance can be quite spotty with the end user subjected to interruption in performance. If we had a combination of WiFi and cellular and could monitor the performance of the user, then if there were problems on WiFI we could hand off to the cellular network – which is much higher performing but has somewhat limited capacity. Where and how to do that is something we are investigating. My guess is the model will depend on specific circumstances: one might imagine either an application on the phone that can manage connectivity, or the network itself could measure the performance of the end user and then take the actions necessary to preserve the QoS they need.”
The other man thrust of Sizer’s thinking is to do with shared access – both in fibre and wireless.
For the former, the company is investigating open access fibre, where the same fibre can be used by a number of different service providers, with each service provider having a different colour separated by wavelengths. Shared access could bring down the “heavy cost” of deploying fibre, opening up the business cases for deploying fibre, which Sizer describes as “still the best fundamental technology to be able to deploy.”
For wireless sharing, Sizer expanded on a prototype Alcatel-Lucent showed at MWC last year – a very high power lightRadio cube – essentially a multi-band small cell that would be able to connect up to four different service providers, with the operators all using the same hardware, backhaul, power and location. A shared small cell would have a big capacity and cost advantage over neutral host DAS, Sizer said.
“Alcatel-Lucent also sells DAS and they have their applications. The problem is it’s still pretty expensive to install and a DAS system only has enough capacity to handle the equivalent of one cellular sector. If you use DAS today you’re going to run out of capacity. The neutral host cube has the ability to serve as many from a single small cell as you could from an entire DAS system, and covers a smaller area so the per-user bandwidth is now dramatically enhanced.”
“That sort of shared access could address the very real costs of deploying high capacity solutions,” Sizer said. “As growth explodes we need to react quite quickly and sharing costs is one area to try to explore.”
Small cells can also play their part, as they are more energy efficient that macro cells
Small cells also form a key part of the thinking of Dr. Thierry Klein PhD, Head of Green Research at Bell Labs. Klien says that with networks now forming 70% of a typical operator’s electricity cost, and with the electric bill for operators becoming a significant item, it is becoming vital that operators can support traffic growth in a sustainable and economically viable way.
Klein is looking at all aspects of global network performance from this point of view, for instance lines of enquiry including how to dynamically turn resources on and off to synch datacentres, so operators don’t need capacity on all the time – making energy usage much more proportional to load.
Small cells, Klein said, can also play their part, as they are more energy efficient that macro cells in delivering signal. “Macro networks could mean you have a mile and a half between the device and the base station, and that’s very inefficient as the wireless medium is not a good propagation mechanism: the receive power is far lower than power transmitted. Shrinking the distance means you gain energy efficiency, so going to small cells is a key direction for improving energy efficiency.”
Coupled to that gain is the ability to power the cells down when you don’t need them, and also to monitor the network and user traffic to dynamically manage small cells.
“We’re looking at the hardware to be able to develop small cells that are low power to start with, and have the capability to monitor power consumption and analyse the traffic in the network to be able to turn the small cell on and off when you need them. That’s a key direction in the wireless area.”