Mukherjee underlined that India has to be more involved in 6G standardization to make sure that its ideas are used globally rather than just locally.
With its vast data volume, growing R&D environment, and developing telecom software and silicon design skills, India is quickly emerging as a key center in the global AI and 6G innovation network, according to US chipmaker Intel. In order to expand its inventions, the business called for more Indian involvement in the development of global 6G standards.
“This is the scale. Udayan Mukherjee, senior fellow and chief architect at Intel’s Network and Communication Group, told Moneycontrol in an interview that “it takes a lot of data to train them, whether it is big language models, smaller AI models, or classical machine learning.”
He went on to say, “The ecosystem being developed in India, for India, is getting significant traction,” alluding to the quick development of industrial capacity and local network architecture.
Intel, which has its biggest engineering and research facility outside of the United States in Bengaluru, is creating software and silicon platforms for next-generation telecommunications networks throughout the world. According to Mukherjee, a large portion of the company’s fundamental design work for platform, packet core, and radio systems currently comes from India.
The influence of India on 6G standards
Mukherjee underlined that India has to be more involved in 6G standardization to make sure that its ideas are used globally rather than just locally.
“You must incorporate all of your discoveries into international standards. Since you lose scale, you do not want to do anything unique that deviates from the norm,” he said. “Even as businesses construct intellectual property on top of international frameworks, India’s inventions must conform to them in order to attain full size.”
In light of India’s renewed determination to take the lead rather than follow in the development of next-generation telecom, he said the country’s current debates and efforts on 6G research, open networks, and spectrum utilization are “steps in the right direction.”
The development of open RAN and Intel’s involvement
Intel has been in the forefront of worldwide deployments in the US, Japan, and Europe, whereas Open RAN (ORAN) is still in its infancy in India. According to Mukherjee, ORAN’s primary advantage is its ability to use virtualization to separate software and hardware, enabling operators to purchase parts from many suppliers rather of being restricted to just one.
“ORAN is all about choice,” he said. “You may get RIC or OSS from one business, radios from another, and distributed and central units from another.” You are only able to work with one vendor, such as Ericsson or Nokia, in conventional RAN. That model is broken by ORAN.
He admitted that while open and disaggregated systems may be more expensive initially, their lifetime economics are better because of their flexibility, competitiveness, and ease of upgrading.
“Although it starts off a little pricier, the price is rapidly declining. Upgrading from 5G to 5G Advanced, for example, is much easier if you are on a virtualized, software-defined infrastructure. Not every advancement requires new hardware.
To assist major network equipment suppliers like Nokia and Ericsson in designing and optimizing solutions on Intel’s platform, Intel supplies silicon, reference software (like FlexRAN), and systems. Additionally, the business works closely with Indian telcos like BSNL, Reliance Jio, and Airtel, either directly or via OEM partners, and produces specialized ASICs for telecom workloads.
Indian telecoms have blamed the sluggish uptake of ORAN on integration costs and ecosystem maturity. Although Mukherjee recognized these difficulties, he said that as standards advance and multi-vendor interoperability becomes better, they are becoming less significant.
“You could get price breaks if you purchase everything from one merchant, but you are stuck with them. Open systems need collaboration between many providers, which initially drives up costs. However, ORAN Alliance requirements are bringing interoperability to reality and lowering costs,” he said.
He went on to say that Intel’s open licensing policy for its reference software promotes creativity among regional producers and Indian system integrators, further reducing reliance on single-vendor networks.
The AI-native network: what makes 6G unique
Considering the future, According to Mukherjee, in order to maximize automation, efficiency, and performance, the 6G network would be both cloud-native and AI-native, meaning intelligence will be integrated at many network tiers.
We are using AI to enhance MIMO performance and channel estimation at Layer 1, the distributed unit. The question of whether AI can beat the finest algorithms is still up for debate, but large-scale testing is taking place,” he added.
According to him, AI will also revolutionize the deployment, management, and upkeep of networks. Machine learning and big language models based on operational data may already improve tasks like fault diagnosis, predictive maintenance, and deployment automation.
Deployment, debugging, and root-cause analysis are examples of day-zero to day-two tasks that AI can automate. To find and address network problems more quickly, you may train models using log data. You do not need to wait for 6G to get this level of operational efficiency,” Mukherjee said.
Monetization is still a problem even as carriers across the globe are making investments in 5G and getting ready for 6G. Although fixed wireless access and data usage are expanding quickly, Mukherjee pointed out that most operators are still having difficulty turning this demand into new sources of income.
“Data and fixed wireless access continue to be the main sources of monetization, notwithstanding all the promises of 5G,” he said. “Although development is gradual, several use cases for private networks and Industry 4.0 are developing.”
According to him, in order to develop new value propositions, networks in the future must become intelligent and compute-embedded. “You can only commercialize data if telecom remains a fat conduit, and that is declining in price.” Intelligent networks that integrate AI, analytics, and computation at the edge provide the true potential.
In order to speed up innovation and commercialization, Mukherjee emphasized the need of ongoing cooperation between telcos, suppliers, the government, and academics.
“Several financing initiatives in the US encourage the development of innovative radio applications. Through programs like BharatNet, India is also taking steps in that direction. It would be essential for BSNL, Jio, Airtel, suppliers, and colleges to work together,” he said.
Mukherjee said that the need for data itself, which is fueled by streaming, artificial intelligence, and new digital applications, continues to drive network development despite present monetization issues. “The need for data is increasing rapidly, if nothing else, due to Netflix, AI, and other factors. That in and of itself keeps suppliers busy and networks crowded. However, he said, “the next phase is integrating intelligence to actually revolutionize how telecom runs.”