The TMN catch-up with Yves Bellego, Director of Technical and Network Strategy, Orange Group, is developing into annual mini-series. This is the third year we have sat down to chat network strategy at MWC
Has much changed over the three years? Well, as Yves himself might say, yes and no. In 2014 the focus was very much on building upon the LTE coverage and capacity the operator had just completed.
VoLTE was proving difficult with a great deal of testing to be done. Small cells were at least a 2016 prospect – although we are in 2016 and still no sign. There wasn’t much love for 5G as a topic, either.
By 2015, thoughts had shifted to the Cloud, to NFV and back to 5G. Bellego was concerned that the NFV ecosystem needed a lot of work. VoLTE was closer but still a work in progress.
So let’s move forward to this year. And one new topic of conversation is the IoT and LoRa. VoLTE seems to be ready, praise be, but is still not launched – although Bellego said it is really now in the hands of marketing rather than the networks/technical team. 5G is still vague in terms of requirements although there is certainly some new system-level thinking.
So here is Bellego, round three. The italics are TMN’s interjections.
LORA is for the short term, EC-GSM and NB-IOT for the long term
One thing about your Orange’s IoT strategy is that Orange has rolled out LoRa in France, but has also committed to NB-IOT (the cellular LPWA technology that is seen as a competitor to LoRa.) What is going on there?
The idea is that in the long term we need radio for internet of things and typically this is one requirement put on 5G. But we cannot wait for 5G and in countries where we cannot even wait for EC-GSM or NB-IOT then we deploy LoRa. If it is a short term need then LORA is the right solution for services to be open in 2016: there is no alternative based on LTE or GSM. That need was clearly identified in France. In other countries we are still assessing if the need is right now or two years from now We prefer solutions that are embedded into our network.
So then what happens to LoRa and those companies selling it. They have a two year window and then they go away? And in a multi-access IoT landscape does the access layer become independent of the service layer, the analytics, all the IoT applications?
The capex cost is done but the opex is pretty low. And this is what we are working on. Typically the idea is to have platforms working on services that can address M2M in 2G, 3G and 4G and IoT that are separate from the radio technology.
VOLTE is ready, IMS is a platform for the future
Over the past three years VoLTE has been a consistent topic of conversation. This year, Bellego said that technology is ready to go.
VoLTE is deployed in almost all our countries. Commercially it is open in Romania and networks are deployed in Spain, France, Belgium. It’s more in the hands of marketing to decide when there will be commercial availability.
In some countries, the introduction of VoLTE has required a concomitant optimisation in the RAN to handle handovers, ensure QOS and so on. Has Orange found that?
We had some cases where voice quality was not very good because of the radio conditions, but that does not meant that optimisation will solve it, but yes there was an impact on voice quality because of radio conditions. One thing is we have CSFB (Circuit Switch FallBack) so it is also about optimising this fallback.
Since we deployed in Romania we have been learning, and there is a bit to optimise on the radio but not only on radio also in the core. In the core it is really a sigificant change.
It’s been hard for operators hasn’t it – a three year project, technically difficult, all to stand still in terms of revenues.
Yes and no. Yes it was a complex project, because we had to update the RAN, the packet core, the IMS and so on. The issue is that touching a service [voice] that has been so fine tuned and optimised is not easy – and the customers are very reactive on voice quality. From the device side there are not so many VoLTE capable devices and that does not help, and the have different configurations.
But that was worth the pain because it is not just about keeping the same service, the same radio but deploying the foundations for the future. By deploying the IMS we have the capability to enhance VoLTE with RCS or any other services based on IMS. So this is the real value of it.
5G – it should enable fixed mobile convergence but it’s unclear how
5G is not just about radio but it also means a new system. That is something we are pushing in research bodies, that research is done on the overall system.
We have been talking convergence fixed and mobile for years but each has developed mobile on their own side -so now we have the opportunity to have a new system.
I have no clue what will be the outcome but for 5G we should not forget the wireline infrastructure that is deployed.
As a fixed and mobile operator, until now in the access each was separate, so what can we do to have more convergence on the access side?We do not have a proprosal but we think if there is a way to have more convergence on access between fibre and 5G, then let’s look at it.
The other thing that we have been trying to push is to have homogeneous performance for the customers. (Bellego describes cell edge fall off in performance in 4G, and the impact of radio conditions on throughput)
It might be that at the end in 5G we have synchronised radio access points – something that we do not have in 4G. These are type of things we want to build.