5G is not only about mobile broadband. It is about creating a new economic value, new business possibilities and an open end-to-end ecosystem to secure people. In terms of 5G readiness, India may not be the first mover in this space, but with the right decisions by telcos, it can come on par with other leading countries. The technology provides lower latency, making various real-time applications possible. In addition, 5G allows handling of billions of devices. This will be important as India enters a mature IoT deployment stage. At a recent event organised by the Cellular Operators Association of India, technology heads of Indian telcos talked about the country’s evolving telecom technology landscape and the key components of a 5G ecosystem. Edited excerpts…
Shyam Prabhakar Mardikar
As the industry has gone digital, there has been an unprecedented increase in both complexity and scale. Interestingly, in India, we have built networks to handle this surge.
With 5G coming in, the product portfolio has become orthogonal, such that the service sets are totally independent of each other. What an enterprise needs is very different from what a machine needs or what a retail customer needs versus what somebody at home needs. Further, there are multiple vectors such as latency, bandwidth, content and security. Therefore, it becomes important to have the ability to understand the scope and span of each one of these service sets at the design stage itself, and then accordingly design a network that delivers a near-optimum level of efficiency, utilisation, topology architecture and management.
With the kind of scale and complexity that 5G will bring in, it will become totally infeasible to manage and run these networks in conventional ways that we are accustomed to. The new technologies require underlying systems to be aware of what is happening at the service plane. These systems are continuously exchanging data with each other, and processing that data so that they can build some kind of intelligence on the fly.
The focus is on creating a service and delivering it to the customer in the best and the most economical way possible. Unless this tenet is built into the end-to-end ecosystem, right from the software layer, which is self-aware, to the infrastructure layer, our ability to run, deploy and create a network of the future will be challenged.
Going forward, in the 5G world, we will not only be talking about a new radio, but also about an entire ecosystem, wherein the hierarchy of the old networks will collapse. In such an ecosystem, every piece of connectivity will just be a single hop from the core to the device.
Role of telcos
In the past six months, the pandemic has given us a big insight into the future as well as the role of telcos. It is imperative to understand that when we move from 3G to 4G, and now to 5G, the service suit is also moving beyond connectivity towards collaboration and productivity. By collaboration, I mean multiple sets of people or industry stakeholders coming together.
Capacity and capability are two key benefits that 5G has to offer. Going forward, the way the world is turning digital, the life of humans, machines and industries will be unimaginable without a digital platform that is supported by 5G.
While the technology landscape is becoming more complex with 5G, several new tools are also now becoming available. The cloud is one technology, which, if orchestrated appropriately, can make service deployment much easier as well as scalable. Further, artificial intelligence (AI) and machine learning (ML) can be used for faster planning and automation as well as for zero-touch installation, self-healing networks, etc. While new technologies make the network very complex, they can also help manage networks and deliver customer experience in a more automated way if deployed at the right time.
At Airtel, we are aware that 5G will bring its own set of complexities. Thus, we are preparing our networks on the core as well as transport side for automated deployments. We have deployed MIMO technology for 5G. We are also working with our infrastructure and network partners to make sites and equipment less power intensive, less space consuming and more efficient to deliver 5G.
AI/ML for business transformation
AI can support functions like planning and predictive capabilities to prepare in advance for any demand surges. Even with ML, one can predict several things. Today, we roll out our networks based on zero touch. The configurations are fed remotely and all the automation happens through a closed-loop process. AI and ML adoption will mature with the advent of 5G.
On the operational front, we are able to predict alarms by using these technologies. There are certain sets of alarms, that give a telltale sign before a hardware goes down. Such signs can be used for predicting which SIM will go down as a SIM does send messages if it is not able to connect to the network. So, if we were to put all this information into algorithms and weave them into predictive network maintenance, we can get a head-start to fix a system before it goes down. We need a lot of tools, trained workforce and effective leadership to establish a digital mindset. It is important to come out of the traditional ways of running networks and start operating in an automated way. One should not be afraid of job loss, but look at more job creation with these new opportunities at the centre.
Going forward, I believe 5G will become an impetus for the government’s Digital India vision by unleashing the power of digitalisation, not only across the telco industry but across verticals such as manufacturing, mining and BFSI.
5G is not like the other G upgrades. It is a completely different architecture – rethought and end to end – including all different domains of the network. We will have a platform that is very much like 5G, even if 5G is not officially adopted. Such a platform will have many capabilities of a 5G network. It is incredibly important to build those capabilities to capture new business opportunities.
I believe we also need to start understanding that the network is now going to behave the way IT set-ups have been behaving for more than 20 years. So we have to look at that network the way it has been looked at in the IT world. This is a very different perspective from the traditional perspective, where we used to buy monolithic systems from network equipment providers and deploy them.
The new capabilities will create several new services, which will be critical for India. India does not need merely cheap connectivity, but it requires telcos to become the base platform, the horizontal industry, for the Digital India programme. For that to happen, telcos need to build strong system integration capabilities, either by themselves or in partnership.
Role of AI/ML
I believe the future is about more automation and augmented intelligence, instead of artificial intelligence. For me, it is a very important distinction because while we will be deploying a lot of AI, we are always going to have human beings and all these technologies will only assist them. We have been doing zero-touch provisioning in our transport networks for a while now.
AI and ML technologies are absolutely essential for running a very large-scale cloud, which will have to be deployed for our networks eventually. The only real way to manage a decoupled software and hardware situation is through the use of AI/ML technologies. In order to do all of that, a telco needs to have a proper understanding of its vision for the future networks it wants to build and deploy.
The point about India being the software factory of the world is incredibly important and critical. As we start to open up our networks, whether they are core networks, or radio or transport networks, software will become more invaluable.
Currently, there are several challenges that India is facing. As we know it, India is a very large country, and thus there are on-the-ground deployment challenges as well as environmental challenges. A software that can address all these challenges can work anywhere in the world. So, I think it is a great opportunity for India, the software industry and the telecom industry to contribute to the rest of the world.
Focus on continuous implementation, continuous deployment
It is also very important to realise that the telecom industry is emerging as the base for a whole lot of other industries. Telecom has been a horizontal set-up traditionally, with telcos responsible for network maintenance and downtimes. Now, with the telecom industry being the base of the digital economy of India, the telcos have to ensure that the network runs and operates without any disruption and gets updated continuously. Thus, continuous implementation, continuous deployment (CICD) is the new norm.
Going forward, the development of CICD capabilities will be the next big thing that the entire industry will be working on. And this requires technologies like AI and ML to be adopted. At the same time, it requires a very different way of thinking about our own networks and how they are operating. We have to start thinking more deeply about the services that we run on top of these networks.
We have already decoupled our networks at Vodafone India (Vi) to a large extent. Our core is completely disaggregated, which means we buy our hardware and software separately and then have it integrated with a third-party system integrator. Currently, we are not doing this on the transport side. Softwarisation will create many more capabilities in our 4G networks, and we essentially do not need to wait for 5G to come up with a full set of slicing capabilities. I strongly believe that there are things we can do with 4G that will start generating revenue very quickly. For that to happen, things like white box routers will have to be adopted. Similarly I believe O-RAN is going to be very important for India, not only to lower costs, but also to create many more capabilities. Once we open up RAN, we can do so much more caching, content provisioning, etc., closer to the edge for the customer. Facial recognition, caching and SaaS become much more possible with edge cloud.
Interestingly, all the things that we are doing might be the first in the telecom world, but it is not the first time in the world. Many of these things have been implemented by hyperscalers in their businesses. So, there are methodologies that we can import from hyperscalers and customise for our businesses to create a world-class universal cloud edge capability on our network.