Telco transformation – Combining all the right ingredients
Mathew Oommen, president of Reliance Jio, in a video posted on the website of Light Reading, shares on how Jio is 5G-ready now, and how next-gen networks will leverage open networking and OpenRAN architectures to take advantage of the benefits of a multi-vendor ecosystem.
Jio has had a tremendous run. Please give us an update on the latest.
As you may be aware, when we launched our services in September 2016, there was about 0.2 exabytes of monthly mobile traffic in India as compared to more than 4 exabytes currently. In the mobile segment, we have been an enabler for digital transformation, and very soon we will be taking this disruption into the enterprise segment as well as the home portfolio.
Now what next? You have made a lot of customers happy, and now there are a lot of high expectations from your company. How are you going to handle the next phase?
In fact, yesterday, I was asked where we are with respect to 5G. If there is one operator in India that is 5G-ready, that is Reliance Jio. This is because ours is the only all-IP, all-VoLTE network in India today.
I believe that is a critical fundamental building block to 5G. Unlike in previous technologies, where 2G, 3G, and 4G were a forklift upgrade, in the case of 5G it shall be a game of coexistence because when you pick up your phone, you are going to see both 4G and 5G on the device. To have the leadership in 5G, the CSP has to be leading in 4G, and one of the key building blocks will be the all-IP VoLTE network.
The second building block is deep fiber that we are continuing to leverage because we believe that fiber is the critical component for successful 5G.
Third is automation as well as what I refer to as intelligent programmable.
Fourth is HetNet, there are different perspectives of what HetNet is. For us, Hetnet is macro, micro, small cells, Wi-Fi as well as FTTx, and this is what makes us ready for continuous growth because we look at our network as a platform, what we call, XG-ready. I use that term because for us, whether it is 5G or 6G, it is pluggable into our network.
Please elaborate on the idea of open networking in general, and what role it plays in your strategy? Also, please comment on O-RAN, another topic of discussion among analysts.
In fact, we are on the board of O-RAN Alliance and have already started the disaggregation process on our network. For us, disaggregation means hardware-software disaggregation as well as control plane disaggregation. So far, there was a sort of vendor lock-in, if I may use that term, but O-RAN gives the opportunity, when talking about the network as a platform, to have pluggable radio units and baseband units from different vendors. So that means, I can technically, over a COTS hardware, have a radio interface as a network function or a container as a function. In other words, O-RAN enables open interfaces, and open methodology, if I may use that term, because even if you have open interfaces, to manage all these interfaces, an open orchestration is needed. Bottom line is, we are accelerating the support, as well as implementing O-RAN on a trial basis, as we speak.
When I think of Linux, which was the best open-source platform, it became successful because someone like Red Hat was actually able to harden it and deliver a deployable product as well as a supportable open-source platform, whereby Linux became scalable.
For telecom transformation, for leveraging hardened open-source platforms, it is important for someone to take those open-source elements and open interfaces, harden it, so that it can become deployable and supportive.
We consider Radisys to be our Red Hat for telco transformation in an open environment, in creating blueprints and reference designs that are hardened and deployable, that either an ODM or an operator can have a plug-and-play module.
Specifically, where is the right layer in your network? Are we talking south side, city level, or regional level? What is the edge for you?
For us, first of all, when we think about the edge, we look at it as a distributed intelligent edge. The idea is even when we think about the intelligence at the edge, we think about intelligence and privacy and trust for the people, privacy and trust for the machines, and privacy and trust for the sensors. Why is that? Because when everything and everyone is getting connected, there is a lot of bandwidth and real-time processing and need for low latency. To meet these criteria, we want to push it as to the edge as possible. In India, we think it can go up to the edge of the cell where the distributed intelligence can indeed happen.
That is a pretty ambitious way to look at it. The biggest buzzwords in technology in general and in networking now are machine learning and artificial intelligence. Please give us a quick summation of how you are looking at that?
We look at AI and ML as a platform, not necessarily as a technology per se, a platform we leverage in multiple layers. The layers will be our operational layer, our customer care layer, and our recommendations-and-offers layer. That platform we are already using in our converged operations. Because we do not have operations for fixed lines, and operations for internet or mobile network, we just have converged operations because we are the only operator that we know of in India that can actually do fixed-mobile convergence. To do that, there is a set of complexity. And the way to manage the complexity in terms of delivering better network performance, better network utilization with enhanced customer experience is, indeed, an AI and ML as a platform. Our first application, so far we have leveraged, is for converged operations.
Talking of edge, we are trying to understand about what goes where in this network?
One of the key criteria for determining in terms of whether it is at the cell side, whether it is at the aggregation side, or at regional datacenter also depends on the types of workloads, because the types of workloads in terms of real-time processing, low latency, type of bandwidth, all of that determines where the right location for intelligent edge computers is.
You could translate that to, from consumer to enterprise, all those levels?
Indeed, that is why today we are focusing on consumers because that is the mobile space that we have entered. In the next few weeks or in the next immediate term you should see us get into the enterprise, as well as the home and we would leverage this platform’s capability across various segments.