The integration of AI at scale is driving tectonic shifts in the way businesses operate, with significant implications for the telecom sector. AI on networks requires workloads to be distributed between devices, edge locations, data centres and the cloud, creating demands for high bandwidth and low latency. AI for networks, where telecom operators deploy AI for traffic forecasting, fault detection, maintenance, network optimisation, spectrum and energy management, etc., creates another set of demands.
In order to meet these requirements, India’s 5G networks must evolve to stand-alone (SA) architectures, which will enable them to deploy advanced 5G capabilities with network slicing, RedCap, lower-latency services and application programming interfaces. This would further enable technologies such as voice over new radio, satellite integration and edge computing.
In terms of physics, data needs to be transmitted in much larger volumes, at higher speeds and with lower latency. This creates a need for greater fiberisation, more efficient use of street furniture, network-ready building designs and smoother right-of-way processes. While Jio and Airtel have been responsible for much of the fibre roll-out, projects such as BharatNet have also made progress in extending connectivity to rural and semi-urban India.
A large part of improving urban connectivity involves developing better coordination among telecom operators, municipal agencies and real estate developers. For example, in-building solutions can allow multiple operators to share the same infrastructure inside a building.
Beyond connectivity, it is important to ensure that AI is deployed in a reliable and sustainable manner. This means paying due attention to power and its sources. Digital infrastructure rests on reliable, sustainable and affordable power. Since India’s AI sector is expected to double every 30 months or so, crossing the $20 billion mark in 2027, this is critical. In high-density computing environments, electricity consumption rises sharply. Power capacity is central to business continuity. Therefore, operators deploying AI across their networks must also plan for resilient power strategies.
Monetising 5G is a puzzle operators have not yet solved entirely satisfactorily. After a fast (but expensive) roll-out that covers most of India, operators need to generate new revenue streams. They must now move to premiumisation of 5G networks, offering enterprise solutions, differentiated quality and speed tiers, content bundles, FWA and portfolios of usage-based services.
The debate around net neutrality has resurfaced in this context. The issue is linked to network slicing, which creates virtual sections of the same physical network with different speeds, latency, reliability and security for different applications or users. The government is now considering revisiting the neutrality policy framework. TRAI has floated a draft, bringing the issue back into focus. The policy challenge will be to allow operators to make full use of these capabilities without undermining open access.
AI, combined with 5G (and with 6G in the future), has improved customer experience, creating AI-driven revenue streams, and improving network planning and operations. AI has also helped with fraud and scam detection, cybersecurity, and operations and maintenance in many ways.
The convergence of the IndiaAI Mission, the Bharat 6G Alliance and the Semiconductor Mission will help India develop connectivity, compute and AI capabilities simultaneously. As India learns to use 5G and AI together, this could provide a big boost to economic activity.