India Mobile Congress 2026: Jio, Airtel, BSNL, Vi and the Race to Build India's Next-Generation Networks
India Mobile Congress 2026 ran from October 7 to 10 at Yashobhoomi in New Delhi, with telecom operators and technology suppliers outlining plans spanning 5G, early 6G research, AI-driven network operations, enterprise cloud and satellite connectivity.
Reliance Jio showcased indigenous radio technology and pre-6G work. Bharti Airtel focused on cloud infrastructure, data centres and managed security. Vodafone Idea presented AI-based customer services and industrial connectivity concepts, while BSNL announced network-security, government-messaging and optical-transport initiatives.
Equipment vendors and industry bodies filled out the programme. Ericsson demonstrated network automation, programmable connectivity and mission-critical use cases. HCLTech focused on the software and systems integration behind telecom operations. MeitY's pavilion covered digital public infrastructure, language technology and semiconductors, while GSMA highlighted standardised network APIs.
Several announcements remain at the prototype, research or planning stage. This report distinguishes those from services already in operation and notes where companies have not released performance data.
IMC 2026 at a glance
Dates: October 7 to 10, 2026
Venue: Yashobhoomi, New Delhi, India
Theme: "Scale Without Boundaries"
Organisers: Department of Telecommunications (DoT) and Cellular Operators Association of India (COAI)
Key areas: 5G and 6G, AI, telecom equipment, semiconductors, satellite connectivity, cybersecurity, enterprise networks, digital public infrastructure and startups
Central question: can India move from deploying connectivity at scale to designing, manufacturing, securing and exporting more of the technology behind it?
The government said the tenth edition drew participants from more than 100 countries, with over 2,000 technology use cases, more than 400 startups and over 900 thought leaders.
In its October 9 summary, the Press Information Bureau framed the event around indigenous technology development and India's role in global telecom. MeitY's pavilion extended that focus to digital public infrastructure, language technology, semiconductors and startups.
India's telecom strategy: from network rollout to technology leadership
India's telecom market has expanded through low-cost data, nationwide 4G adoption and a rapid 5G rollout. At IMC 2026, government officials put more emphasis on domestic design and manufacturing of the equipment and software used to build networks.
Communications Minister Jyotiraditya M. Scindia said the next decade should focus on designing, developing and manufacturing technology in India, for domestic use and for export. The policy places telecom alongside other strategic technology industries.
A network can be deployed in India while relying on foreign-designed radio equipment, chips, software, cloud services and specialist components. Local manufacturing helps, but real capability also covers system design, software development, intellectual property, standards participation, security testing and long-term maintenance.
A complete domestic ecosystem has to cover several layers:
Radio access: antennas, radio units, baseband processing, timing and synchronisation.
Transport: fibre, optical transmission, routing and backhaul.
Core network: subscriber management, authentication, session management, policy control and service orchestration.
Cloud and compute: infrastructure for network functions, AI applications and enterprise workloads.
Software and operations: automation, observability, billing, service assurance and customer systems.
Security: identity, access control, supply-chain assurance, threat detection and recovery.
Standards and intellectual property: contributions that can shape interoperable global products.
Reliance Jio: indigenous 5G technology, pre-6G research and AI-first connectivity
Reliance Jio built its showcase around domestic technology development. The operator said it is developing proprietary pre-6G technology and displayed a Giga MIMO Radio next to a common baseband unit meant to support both 5G and future 6G evolution. Jio's leadership also argued that India could be among the first countries to deploy 6G once global standards are finalised.
ETTelecom reported the announcements on October 9; they remain part of Jio's development roadmap. A pre-6G demonstration is not a final 6G standard, and a common baseband platform does not prove that future radio equipment will interoperate across vendors.
What a Giga MIMO radio could mean
MIMO stands for multiple-input, multiple-output. Modern cellular systems use multiple antennas at the transmitter and receiver to raise spectral efficiency and serve several users at once. Massive MIMO pushes this further with larger antenna arrays and heavier signal processing.
Jio did not publish detailed performance figures for the Giga MIMO Radio in the event material reviewed. Its practical value will depend on supported spectrum bands, antenna design, throughput under load, power consumption and integration with the operator's existing radio network.
A common baseband platform for 5G and 6G evolution
The baseband processes digital signals and runs key parts of the radio protocol stack, including modulation, coding, scheduling and the handling of radio traffic. A platform designed to span generations could cut duplication and make it easier to reuse hardware and software as standards develop.
Because 6G specifications are still under development, the extent to which the baseband can be reused will depend on the final radio interface, spectrum and architecture adopted by standards bodies.
Jio's 6G ambition and the standards race
According to the ETTelecom report, Jio said it has filed more than 8,000 patents, including hundreds related to 6G, and submitted over 934 technical contributions to standardisation bodies. These are company-reported figures. A patent count does not tell you how much influence a company will have over a future standard.
Patent totals and standards submissions indicate the scale of Jio's activity, but do not show how many proposals will be adopted in 6G specifications.
Jio's reported aim of placing India among early 6G adopters is a long-term goal. Getting from research to commercial service depends on global standards, spectrum policy, devices, network equipment, testing and a clear business case.
AI-native and quantum-safe networking
Jio also described work on AI-native and quantum-safe network architectures, which address different problems.
An AI-native network uses machine-learning systems in tasks such as traffic forecasting, fault detection and radio-resource management. Jio did not detail which functions are already running in production or how much of the network can be managed autonomously.
Jio also referred to quantum-safe networking. In practice, operators are assessing post-quantum cryptography for long-lived network equipment, certificates and management systems. The company did not disclose a deployment timeline or the specific technologies included in its demonstration.
AI use cases beyond the smartphone
Jio also pointed to agriculture, healthcare and education as potential beneficiaries of AI-enabled connectivity, although it did not announce specific deployments or customer results for those use cases at IMC.
Bharti Airtel: sovereign cloud, data centres and managed cybersecurity
Bharti Airtel's IMC 2026 presence centred on enterprise infrastructure instead of one consumer device. Chairman Sunil Bharti Mittal demonstrated Airtel Cloud's sovereign capabilities, its multi-availability-zone architecture, Nxtra data-centre operations and the company's AI-powered intelligent Security Operations Centre, called iSoC. The company also highlighted research into quantum-powered network optimisation with the Bharti School of Telecommunication Technology and Management at IIT Delhi. ETTelecom covered the announcements on October 9.
What a sovereign cloud is designed to address
A sovereign cloud is designed to keep data and cloud operations within a specified legal jurisdiction. Airtel positioned its cloud offering for customers that need Indian data residency and domestic infrastructure.
Airtel Cloud's showcase stressed domestic data residency and multi-availability-zone capability for mission-critical workloads. A multi-availability-zone design can keep workloads running if one site fails, provided applications replicate data and fail over correctly. Airtel did not publish availability figures or recovery-time targets, or detail all controls covering backups, support access and third-party services.
Nxtra and the energy cost of digital infrastructure
Airtel also highlighted Nxtra's data-centre capabilities and the use of AI to optimise energy efficiency. Data centres draw power for servers, storage, networking and cooling. As AI workloads grow, electricity supply, cooling design and utilisation become the main operating constraints.
Airtel said AI would be used to improve data-centre energy efficiency. It did not disclose performance figures such as power usage effectiveness or energy consumed per unit of compute.
Airtel iSoC: managed security for enterprises
Airtel showed its AI-powered intelligent Security Operations Centre, described as a five-layer security offering with 24/7 managed defence. According to ETTelecom, the company said it is already deployed for enterprise clients including Indian Railways and the Employees' Provident Fund Organisation (EPFO).
A managed security operations centre collects and correlates telemetry, investigates alerts and supports incident response for customers. Airtel described iSoC as a five-layer service with 24-hour coverage.
The service combines monitoring and incident response across customer environments. Airtel has not published independent performance data on detection rates, false positives or incident-resolution times.
Quantum-inspired optimisation research
Airtel highlighted research with IIT Delhi into quantum-powered optimisation of telecom network routes. Routing is a hard optimisation problem: operators move traffic across fibre and transport networks while balancing capacity, latency, reliability, power consumption and the risk of congestion.
Airtel presented the work as research with IIT Delhi. No benchmark results or deployment timeline were disclosed, so it remains unclear how the approach compares with conventional network-routing algorithms.
Airtel's postpaid changes and the business model behind network investment
Alongside the technology showcase, Airtel introduced revised postpaid plans from October 8, 2026, with a ₹50 increase across the affected tiers and a complimentary annual international roaming trip, as reported by ETTelecom.
Airtel's revised postpaid plans took effect on October 8, with a ₹50 increase across the affected tiers and an annual international roaming trip included, according to ETTelecom.
Vodafone Idea: Vinzy, smart mining, connected mobility and AI-assisted enterprise services
Vodafone Idea came to IMC 2026 with practical services for consumers and enterprises: a personal AI assistant, zero-touch voice-based customer service, smart mining, connected mobility and AI-powered tools for small and medium-sized businesses.
ETTelecom outlined Vi's planned demonstrations on October 8. TelecomTalk separately reported that Vinzy was shown as a prototype meant to run inside the Vi app, and that the company had not confirmed whether or when it would launch commercially.
Vinzy: a personal AI assistant for voice interactions
Vinzy was presented as an AI assistant that can handle calls in preferred regional languages. The idea is to let customers talk to telecom services in natural language instead of working through menus or reading long instructions.
A voice assistant might screen calls, summarise messages, answer common questions or route a customer to the right service. Regional-language support could open the service to people who are less comfortable with English-language interfaces.
Vinzy's performance will depend on how accurately it handles regional accents, code-switching and noisy calls. Vi has not announced launch timing or published details on audio retention and processing.
Zero-touch voice-based customer service
Vi also planned to demonstrate AI-supported voice service for customer support. Older automated systems rely on keypad menus or fixed speech trees. More flexible conversational systems can interpret a request, ask clarifying questions and pull up the relevant account or troubleshooting information.
For the service to resolve account issues rather than simply answer questions, it must connect to Vi's billing and customer-management systems. Vi has not detailed the backend integrations planned for the demonstration.
Smart mining: private 5G, edge compute and connected wearables
Vi's industrial showcase combined private networks, edge computing, AI analytics and connected wearables for mining. Mines are a demanding place for connectivity: tunnels and underground workings limit radio propagation, GPS is generally unavailable below ground, and workers face gas, heat, machinery and structural instability.
A private cellular network gives managed connectivity across a site, and edge computing lets selected analytics run close to the equipment. Wearables and sensors can send telemetry such as worker location within the supported system, environmental measurements or equipment status. AI can flag patterns worth investigating, but safety-critical alerts need predictable performance and fail-safe behaviour.
Vi did not publish field-test results or reliability figures for the mining demonstration.
Connected mobility: eCall, ADAS, V2X and eSIM management
Vi also highlighted connected-mobility capabilities for automotive manufacturers: eCall, advanced driver-assistance systems (ADAS), vehicle-to-everything (V2X) communication, in-cabin voice control and eSIM management. They cover different layers of connected transport.
eCall: systems that can send emergency information after a serious road incident, depending on the vehicle and service implementation.
ADAS connectivity: data services that can support driver-assistance features, traffic awareness or remote diagnostics. It complements the vehicle's onboard sensing and safety systems.
V2X: communication between a vehicle and other vehicles, road infrastructure, pedestrians or networks. Its usefulness depends on compatible systems, latency, coverage and deployment density.
In-cabin voice control: natural-language interfaces for selected vehicle functions and connected services.
eSIM management: remote provisioning and lifecycle management of cellular subscriptions embedded in vehicles.
Connected vehicles remain in service for years, often longer than individual mobile-network or software generations. Remote diagnostics and eSIM management therefore depend on secure update systems, device identity controls and long-term vulnerability support. Poorly secured telematics units, mobile apps or backend APIs can also expose vehicle systems to remote attacks.
ReadyForNext and digital tools for MSMEs
Vi also planned to show AI-powered insights and interactive tools on its ReadyForNext digital advisory platform for micro, small and medium-sized enterprises (MSMEs). Small businesses often have no dedicated IT team to assess cloud services, security, digital marketing and operational software.
Vi said ReadyForNext would offer AI-assisted insights and interactive guidance to small businesses. The platform's practical value will depend on the quality of its recommendations and whether businesses can act on them without buying services they do not need.
BSNL: customer service, core-network security, Bharat Sandesh and photonic transport
BSNL announced four initiatives at IMC 2026, covering customer support, cybersecurity, government communications and optical networking. ETTelecom reported them on October 9.
The announcements covered four distinct areas: customer support, network security, government communications and optical transport.
Vaani: multilingual support and network visibility
BSNL said it has updated its Vaani chatbot app with automated location detection and live network-performance views in the customer's preferred regional language. It also described a voice-based complaint process.
Vaani's location and network views are intended to give support teams more context when customers report service problems. Signal strength alone cannot identify every fault: congestion, indoor coverage, Wi-Fi settings, backhaul, DNS and application servers can all affect the experience. The company did not publish technical details on the diagnostic data used or its location-data retention policy.
AI and machine learning at the network layer
BSNL announced a partnership with Indian deep-tech firm Saptang Labs to integrate AI and machine-learning threat protection into the network core. The company described a system that checks traffic against threat intelligence and can act on malicious activity as it happens.
Network-level detection can identify traffic patterns that are difficult to see from a single device, although visibility depends on encryption, network architecture and where inspection takes place. The companies have not published independent detection or false-positive rates for the proposed system. Those figures will be important in assessing how it performs at scale.
Bharat Sandesh: sovereign two-way government messaging
BSNL also unveiled Bharat Sandesh, a two-way messaging platform developed with ViH Messenger. It is meant to let government departments send verified, interactive communications to citizens through domestic infrastructure, with regional-language engagement and support for text, voice and AI-assisted responses.
Bharat Sandesh is designed for two-way communication between government departments and citizens, including regional-language interactions. Its adoption will depend on departmental integration, message authentication and the handling of replies. BSNL and ViH Messenger did not provide deployment figures or independent security-test results in the material reviewed for this article.
Fujitsu and the All-Photonics Network proposal
BSNL signed an expression of intent with Fujitsu Japan to explore an All-Photonics Network (APN) over BSNL's existing dark-fibre infrastructure. The stated goal includes enabling commercial services for hyperscaler and enterprise customers.
Optical networks carry data as light through fibre. Conventional networks often convert signals between optical and electrical form at intermediate stages for switching, routing or regeneration. An all-photonic approach keeps more of the transport path in the optical domain, which removes some conversion steps and can improve energy efficiency for suitable traffic patterns.
Results will depend on the implementation. Not every packet can pass through a transparent optical path without electronic processing, and modern networks still need routing, traffic engineering, monitoring, protection and service control. Optical bypass pays off where topology, traffic demand and service requirements make it practical.
Existing dark fibre could let BSNL evaluate new optical services without laying new routes everywhere. BSNL and Fujitsu will need to identify routes where the architecture improves latency, power use or capacity enough to justify the optical equipment, management systems and integration costs.
Ericsson: AI-enabled networks, differentiated connectivity and Make in India
Ericsson's IMC 2026 showcase linked network technology to enterprise applications, startup innovation, manufacturing and the longer path to 6G. Its official October 7 announcement described demonstrations across network automation, monetisation, mission-critical connectivity, future networks and the Ericsson Innovation Challenge.
AI in the radio access network
Ericsson demonstrated AI-enabled radio technology and network operations, including AI-powered RAN applications, or rApps, for deployment, optimisation, assurance and self-healing.
The radio access network (RAN) connects user devices to the operator's core and has to respond to shifting traffic, interference, mobility and local radio conditions. Automation can adjust parameters, predict congestion, identify faults and cut repetitive manual work.
A rApp analyses network data and recommends or triggers actions within its permitted scope. How useful it is depends on data quality, model accuracy, how often decisions must be made, and whether an action is safe in the current context. Some changes can be automated at low risk. Others need human approval or staged deployment.
Autonomous operations need observability and controlled authority. Operators have to know which data influenced a decision, what configuration changed, whether performance improved and how to roll the change back. A system that cannot explain or reverse a harmful action is hard to trust in a national network.
5G Standalone and differentiated connectivity
Ericsson highlighted 5G Standalone (SA), network slicing, enhanced uplink and service-level guarantees as building blocks for differentiated connectivity. Non-standalone 5G relies on a 4G core for some functions. 5G SA uses a 5G core and supports a more flexible service framework.
Network slicing creates logically separated network environments with different policies and performance targets. It does not create unlimited capacity, because slices still share physical resources. Operators must engineer admission control, radio scheduling, transport capacity and service assurance to meet their commitments.
Differentiated connectivity could let an operator give a business or application a more predictable experience for a particular task, such as a live event, an industrial process or a connected service. The commercial model has to say what performance is guaranteed, under which conditions, how it is measured and what happens when the service misses the commitment.
Network APIs: Quality on Demand and SIM Swap checks
Ericsson also demonstrated programmable network capabilities, including Quality on Demand (QoD) and SIM Swap APIs, part of a wider industry effort to expose selected network functions through standardised interfaces.
A QoD API lets an authorised application request a particular connectivity treatment for a session or use case, subject to operator policy and available network capability. A SIM Swap API helps an authorised service learn whether a SIM change happened recently, which can feed risk scoring for banking or account recovery.
Treat these APIs as extra signals and controls. They are not complete fraud-prevention systems. A recent SIM change raises risk without proving fraud, and the absence of one does not make a transaction safe. Applications still need strong authentication, transaction monitoring, device and account risk checks, and recovery processes that protect users.
Standardisation matters because developers do not want a separate integration for every operator. GSMA Open Gateway and the CAMARA project are working toward more consistent network APIs. Its value depends on how widely the APIs are implemented, the quality of developer tooling, privacy controls and the reliability of the responses.
Smart glasses, stadiums and consumer experiences
Ericsson also paired AI-enabled smart glasses with 5G SA and network APIs in a stadium setting. The concept covered personalised content, navigation, virtual shopping and interactive experiences.
Wearables show why uplink performance and latency matter. A device may capture images or video, receive information from a cloud service and display a response while the wearer moves. Depending on the design, some processing runs on the device and heavier tasks use edge or central cloud infrastructure.
A stadium is a good stress test, since thousands of people are packed into a small area and connectivity has to carry ticketing, payment, streaming, navigation and staff communications at once. A demonstration cannot show how the service will perform for every user at a sold-out event. Capacity planning, interference management and fallback behaviour remain essential.
Privacy matters here too. Smart glasses can capture video, location and context about bystanders who never opted in. A responsible deployment needs clear indicators, user controls and limits on what is captured or transmitted.
Mission-critical connectivity and railway safety
Ericsson highlighted mission-critical connectivity, including a demonstration related to 5G connectivity for India's indigenous KAVACH automatic train protection system, and drone-assisted connectivity for places where ground infrastructure is unavailable or damaged.
Ericsson's KAVACH demonstration concerned the use of 5G connectivity in railway operations. Cellular connectivity would form one part of the system; the safety functions and certification requirements of KAVACH remain separate.
A tethered drone carrying a radio and antenna can restore or extend coverage in emergencies, disaster response or hard-to-reach areas. Its usefulness depends on flight endurance, weather, tether power, airspace rules, backhaul and the security of the temporary network. It is a specialised coverage tool and does not replace permanent infrastructure.
Ericsson Innovation Challenge: turning 5G into applications
Ericsson said six winning startups from its Innovation Challenge were showing solutions that apply 5G and AI to agriculture, healthcare, digital inclusion, advanced manufacturing and environmental sustainability. The official announcement places these projects alongside the company's network demonstrations.
Infrastructure investment is worth more once developers build services that solve real problems. A connected-agriculture project has to do more than send sensor data. It should help a farmer improve yield, reduce waste or save time. A healthcare solution must fit clinical workflows and protect patient data. A manufacturing system must integrate with existing equipment and show measurable operational benefits.
Startups often need test networks, technical mentoring, customers and deployment partners. An exhibition gives visibility, and the harder work comes afterwards: pilots, procurement, compliance, user training and support.
Made in India manufacturing and export ambitions
Ericsson said it has produced more than 15,000 antennas and 25 million components in India since July 2025 and discussed plans to export India-made products. The Indian Express reported the figures.
Ericsson said it has manufactured more than 15,000 antennas and 25 million components in India since July 2025 and discussed exporting products made in the country. Export orders would test whether local production can meet overseas requirements for price, quality and long-term support.
Ericsson also highlighted its Indian research and development capabilities, including local semiconductor talent contributing to its Ericsson Silicon platform. Semiconductor development spans architecture, verification, fabrication coordination, packaging, validation and software integration. The resulting components must perform reliably inside radio equipment and remain supportable over long product lifecycles.
HCLTech: the software and engineering layer behind AI-native telecom
HCLTech's showcase covered a less visible part of telecom transformation: the software, engineering and operational systems that let networks be built, monitored and monetised. Its announced focus areas were autonomous networks, operations and business support systems modernisation, network monetisation, network cloudification, AI-based service assurance, Open RAN engineering and 6G preparation. ETTelecom's October 6 report outlined the company's plans ahead of IMC.
Why OSS and BSS still matter
Operations Support Systems (OSS) help operators manage and monitor network resources, services, faults and performance. Business Support Systems (BSS) handle customer-facing processes such as ordering, charging, billing and account management. Both tend to be large, interconnected and hard to modernise without disrupting service.
As networks become more programmable, the operations layer has to keep up. A network API request, for example, may need to connect a developer portal with identity checks, policy controls, service orchestration, charging and performance monitoring. If those systems do not integrate cleanly, a capable network is still hard to sell or run.
AI can summarise alarms, correlate events, suggest root causes and automate repetitive workflows, as long as it works from reliable data and controlled permissions. A wrong diagnosis sends engineers in the wrong direction, and an automated action can make a fault worse if the system does not understand dependencies.
Service assurance and autonomous operations
Service assurance checks that the network delivers the performance and availability customers expect. In a complex 5G environment, failures can occur in the radio, transport, core, cloud or application layer. A user can see a slow application while the radio link looks healthy.
AI-assisted assurance can correlate events across those layers and point to likely causes. How useful it is depends on telemetry coverage, accurate topology information and the ability to tell correlation from causation. Operators should check whether automation shortens mean time to detect and repair incidents, since the number of alerts processed says little.
Autonomous operations are best introduced step by step. Low-risk recommendations come first, then narrowly scoped actions with clear limits. Consequential changes should need approval, staged rollout and rollback. Human teams must still understand the system well enough to step in when its assumptions fail.
Cloudification and Open RAN engineering
Cloudification runs more network functions on software-based infrastructure instead of specialised fixed-purpose appliances. It makes capacity more flexible and aligns telecom operations with cloud automation practice. It also creates dependencies on compute scheduling, virtualisation, Kubernetes-style orchestration, storage, timing and infrastructure security.
Open RAN engineering adds an integration challenge. Separating radio and baseband functions through defined interfaces broadens supplier choice and can also make system performance harder to predict. Operators need consistent timing, fronthaul capacity, software versions, security patches and end-to-end testing.
HCLTech's role shows that telecom transformation involves more than new radios. Operators need systems integrators and software teams who can validate how components work together in a production network.
MeitY's pavilion: digital public infrastructure, language AI and semiconductors
MeitY's pavilion gathered services and capabilities that underpin India's wider digital ecosystem. The ministry said its showcase included Aadhaar, DigiLocker, UMANG, Entity Locker, API Setu, UX4G, Digital India BHASHINI, MyGov, C-DAC, NIXI, SAMEER, India Semiconductor Mission and startups incubated through Software Technology Parks of India (STPI). The official overview is available from the Press Information Bureau.
Digital public infrastructure as a connectivity use case
Aadhaar, DigiLocker, UMANG and Entity Locker show how identity, documents and public services can be delivered digitally. They are not telecom technologies, yet their reliability depends on secure networks, interoperable APIs, usable devices and trusted authentication.
Digital public services depend on more than a working mobile connection. Failures in hosting, identity verification or APIs can interrupt a transaction even when the network is available.
BHASHINI and Indian-language access
Digital India BHASHINI demonstrated AI and language technologies meant to make digital content and services accessible across Indian languages. Language support is a practical problem for telecom and public digital services alike. An interface that works in only a few languages shuts some users out or forces them to rely on intermediaries.
Speech recognition, machine translation and text-to-speech can lower these barriers, but quality varies by language, dialect, recording conditions and domain. A translation that works for general browsing may fall short for a medical instruction, a legal notice or a financial transaction. Systems should signal uncertainty, support human review for high-stakes use and avoid treating model output as authoritative.
API Setu and interoperability
API Setu's presence highlighted how application programming interfaces connect digital services. APIs let systems exchange data or request functions under defined rules. They make integration easier, and each one is also a potential security boundary.
Secure API design requires authentication, authorisation, input validation, rate limiting, monitoring, version management and a clear data-access policy. Developers should not assume an API is safe because it belongs to a government or enterprise platform. Permissions should be scoped to the minimum needed, and sensitive actions should leave auditable records.
Telecom faces the same trade-off as digital public infrastructure: more integration creates more value and raises the importance of identity and consistent security controls.
C-DAC, NIXI, SAMEER and the research ecosystem
The pavilion included the Centre for Development of Advanced Computing (C-DAC), National Internet Exchange of India (NIXI), Society for Applied Microwave Electronics Engineering & Research (SAMEER) and India Semiconductor Mission (ISM).
Each sits at a different point in the stack. Advanced computing supports high-performance systems and software research. Internet exchange infrastructure helps networks swap traffic efficiently. Microwave and electronics research feeds radio-frequency and communications engineering. Semiconductor initiatives address a strategically important part of the hardware ecosystem.
Their presence is a reminder that a telecom network depends on domestic research, Internet routing, compute, testing, components and skilled people, on top of the mobile operator and the radio vendor.
Startups and the route from prototype to deployment
According to the ministry's announcement, STPI brought 14 incubated startups to the event. Startup exhibitions can connect developers with customers, investors and mentors. Moving from demo to product still takes a clear buyer, an integration plan, support arrangements, security review and evidence that the technology beats existing alternatives.
GSMA, Open Gateway and the move toward standardised network APIs
GSMA's pavilion at IMC 2026 highlighted Open Gateway and GSMA Fusion, with a focus on standardised network APIs. It also covered digital trust and anti-scam collaboration, and presented the Tokyo Accord as a shared industry vision for a less fragmented transition to AI-native 6G. Event information is on the GSMA's IMC 2026 page.
Why network APIs matter
Operators control capabilities that applications want: connectivity quality, device and SIM status, location-related signals and other network functions. Developers have historically had to integrate with each operator separately or work through complex carrier-specific systems.
Standardised APIs aim to expose selected capabilities through more consistent interfaces. That can cut integration costs and open opportunities in fraud prevention, connected devices, event experiences and enterprise services.
Open Gateway's appeal depends on operators implementing APIs consistently. Developers will also need usable test environments, predictable error handling, transparent pricing and clear rules for consent and data protection.
SIM Swap APIs and fraud risk
A SIM Swap API lets an authorised service check whether a subscriber's SIM changed recently. Banks and other services can use the result as one input to fraud screening; it does not establish by itself whether a transaction is fraudulent.
Quality on Demand and enterprise services
Quality on Demand APIs let an application request a particular level of network treatment where the operator supports it. That could help live video, industrial control or interactive experiences, provided the requested service maps to real network resources and policies.
Quality on Demand APIs let applications request a defined level of connectivity where the operator supports it. Availability and performance remain subject to the operator's network capacity and service terms.
Digital trust and the anti-scam ecosystem
GSMA's trust showcase focused on collaborative ways to protect identities and digital interactions. Scams routinely cross institutional boundaries: a fraudulent message may start on one platform, send the victim to a fake website and end at a bank account or digital wallet.
Satellite connectivity: Starlink, terrestrial operators and the spectrum debate
Satellite broadband stayed a major policy issue around IMC 2026. Starlink representatives said the company has set up around 20 gateway sites in India and described India-specific security and data-localisation measures. The Economic Times and Mint reported that the company was still waiting for final regulatory and security clearances.
Building gateways does not authorise commercial service. A satellite provider can invest in infrastructure and still need approvals on licensing, spectrum, security, lawful interception, ground infrastructure and operational compliance.
Where satellite broadband can help
Low Earth orbit (LEO) satellites operate closer to Earth than geostationary satellites, reducing signal travel time. Service performance still depends on constellation capacity, terminal hardware, gateway placement, spectrum and congestion. Satellite broadband is likely to serve areas where terrestrial networks are unavailable or uneconomic, rather than replace fibre and mobile networks in dense markets.
The gateway and data-routing questions
Gateways connect satellite networks to terrestrial infrastructure. Their location and routing architecture affect latency, resilience, regulatory oversight and how traffic is handled. Data-localisation claims need a detailed look: where user traffic is routed, where metadata is processed, how support access is managed and which entities can reach logs or operational systems.
Spectrum allocation and competitive neutrality
Terrestrial operators such as Airtel and Jio have argued that satellite services competing in retail broadband or mobility markets should face a fair regulatory framework. Satellite providers respond that their networks have different technical and economic characteristics. The policy task is to assign spectrum and licence conditions in a way that supports connectivity, manages interference, protects national security and keeps competition workable.
Spectrum assignment, licensing and final operational approval are separate issues, and a company can hold one authorisation while still needing another. The public should look for explicit government approvals and a formal commercial launch announcement. A conference demonstration does not mean a service is available.
Direct-to-device satellite services
Direct-to-device (D2D) connectivity lets compatible phones communicate through satellites for selected services, potentially including messaging or emergency connectivity. It is technically different from a dedicated satellite broadband terminal. Device compatibility, spectrum arrangements, antenna constraints, satellite capacity and integration with terrestrial mobile networks all matter.
D2D can be valuable where there is no cellular coverage, though it is not necessarily equivalent to full broadband. Early services may have narrower use cases, limited bandwidth or restrictions by device and location. Regulatory decisions will shape which services are offered and under what conditions.
Smartphone and edge AI: local compute joins the telecom story
IMC 2026 also had consumer-device news. Event reporting said Motorola showed the Motorola Signature 27 and iQOO confirmed the upcoming iQOO 16, both tied to Qualcomm's Snapdragon 8 Elite Extreme Gen 6 platform. Gadgets 360 has the coverage.
AI models are also moving onto smartphones and other edge devices. Local processing cuts round-trip latency, limits how much personal data goes to cloud services and lets some features work without a continuous connection.
Why on-device AI is useful
A phone with a capable neural processing unit (NPU) can run tasks such as speech recognition, image enhancement, summarisation and small language-model inference. How useful it is depends on the model and workload. A compact model is fast and efficient but less capable than a larger cloud model. A hybrid system handles simple tasks locally and sends complex requests to remote infrastructure.
On-device AI can reduce reliance on cloud processing, although individual applications may still upload prompts, images or telemetry. The network impact will depend on how manufacturers divide processing between devices, edge servers and central cloud systems.
Thermal and power limits
Smartphones have strict power and cooling limits. Sustained AI workloads raise temperature, drain the battery and trigger throttling. A chipset's advertised AI capability does not predict how fast it will run a workload after several minutes of continuous use.
Independent testing should measure response time, power draw, sustained performance, memory use and result quality. A manufacturing-process label alone does not prove one device will outperform another in every scenario.
Edge AI changes network demand
If a device processes more locally, it may send less raw data to the cloud. Edge AI can also encourage richer applications that exchange more images, video and contextual data, so the net effect on traffic depends on application design.
Operators therefore have to plan for a mixed architecture: some inference on devices, some at nearby edge locations and some in central cloud environments. The right split depends on latency, privacy, model size, energy use, cost and what happens when a connection fails.
Engineering challenges behind AI-native telecom
Across operator and vendor showcases, AI turned up in customer service, radio optimisation, security monitoring, enterprise applications and consumer devices. The demonstrations covered several distinct uses of AI, from traffic forecasting to automated network operations.
Predictive analytics
Operators are exploring predictive models for traffic forecasting, fault detection and maintenance planning. Their accuracy depends on the quality of network telemetry and how well models adapt to changing traffic patterns.
Generative AI for operations
Generative AI tools can search technical documentation, summarise alarms and assist with incident reports. Operators are also testing how to connect these tools to live telemetry without allowing unverified output to trigger network changes.
Agentic workflows
Agent-based systems could automate multi-step tasks such as collecting diagnostics and comparing network metrics. Telecom operators have not yet provided enough deployment data to assess how broadly such systems are being used in production networks.
AI security and model risks
AI systems bring their own security questions: prompt injection, data leakage, model misuse, poisoned training data, insecure tool access and excessive privileges. An operator that uses AI to manage network functions has to secure the model interface and every tool the model can call. Network security does not reduce to feeding traffic into a model.
Measuring success
The clearest measures of AI's value in telecom will be changes in outage duration, fault-resolution time, energy use, network performance and operating cost.
Open RAN and the question of vendor flexibility
Open RAN separates parts of the radio access network through defined interfaces so components from different suppliers can work together. That gives operators more choice and encourages innovation in radio hardware, software and automation.
Potential benefits include less dependence on a single supplier, more flexible upgrade paths and openings for domestic equipment and software companies. Openness does not guarantee easy interoperability, because implementations vary in performance, timing behaviour, management, security and feature support.
A production deployment needs testing across the radio unit, distributed unit, central unit, transport network and management system. Operators must also plan for software updates, security patches, fault isolation and who is responsible when a multi-vendor system fails. Integration costs can eat into the savings from supplier competition.
Optical transport and fibre: the foundation under 5G and 6G
Fibre backhaul links radio sites to aggregation points, core networks and data centres. As mobile traffic grows, operators are expanding transport capacity to limit congestion and support lower-latency services.
BSNL's proposed APN work with Fujitsu is one approach to better optical transport, and other operators and vendors keep investing in fibre, routing and more efficient architectures. The design has to match the service requirement.
Fibre construction remains expensive, particularly where civil works are difficult. Existing dark fibre may support new services, while satellite and microwave links remain options where fibre is impractical.
Telecom security: a larger software attack surface
Operators are extending software-based functions across radio access, core networks, cloud infrastructure and enterprise APIs. That creates more interfaces to secure and more dependencies to track across suppliers and managed services.
At IMC, BSNL's planned network-core protection with Saptang Labs and Airtel's iSoC were the clearest security-focused announcements. Neither company supplied independent detection or false-positive figures in the material reviewed. Those results, alongside deployment scope and incident-response performance, will be needed to assess the services beyond their exhibition claims.
What the announcements mean for India's telecom market
For operators, the exhibits pointed to continued investment in radio equipment, fibre, cloud infrastructure and automation as traffic and enterprise requirements change. For enterprise customers, private networks, managed security and network APIs are moving further into the operator portfolio.
The satellite debate remains unresolved. Starlink's reported gateway infrastructure is not the same as permission to launch a commercial service, and spectrum and security approvals remain separate parts of the process.
What to watch after IMC 2026
The next milestones are commercial launches and operator deployments. Jio's pre-6G work remains in development; Vi has not confirmed a launch date for Vinzy; and BSNL and Fujitsu have signed an expression of intent to explore an All-Photonics Network. Wider adoption of standardised network APIs will depend on operator support.
Frequently asked questions
What was the main focus of India Mobile Congress 2026?
IMC 2026 focused on 5G and 6G, AI, telecom equipment and manufacturing, satellite connectivity, cybersecurity, enterprise networks, digital public infrastructure and startups. It took place at Yashobhoomi in New Delhi from October 7 to 10, 2026.
What did Reliance Jio showcase at IMC 2026?
Jio highlighted indigenous telecom technology, including its pre-6G stack, Giga MIMO Radio and a common baseband unit meant to support 5G and future 6G evolution. It also discussed AI-native and quantum-safe network research. These announcements reflect technology development and long-term ambitions. They are not a commercial 6G launch.
What did Airtel showcase at IMC 2026?
Airtel highlighted Airtel Cloud's sovereign capabilities, multi-availability-zone infrastructure, Nxtra data centres, its AI-powered intelligent Security Operations Centre, and research with IIT Delhi into quantum-powered network optimisation.
What did Vodafone Idea bring to IMC 2026?
Vi announced a showcase covering Vinzy, an AI assistant presented as a prototype, voice-based customer support, smart mining using private networks and edge computing, connected mobility capabilities and digital advisory tools for MSMEs.
What did BSNL announce at IMC 2026?
BSNL announced updates to its Vaani customer-support app, a partnership with Saptang Labs for AI-enabled network security, Bharat Sandesh for sovereign government messaging, and an expression of intent with Fujitsu to explore an All-Photonics Network over BSNL's dark fibre.
What did Ericsson demonstrate?
Ericsson showed AI-enabled network operations, radio access automation, 5G Standalone and differentiated connectivity, network APIs, mission-critical connectivity, 6G-related concepts, startup innovations and its Indian manufacturing and R&D capabilities.
What is the difference between 5G and 6G?
5G is a deployed generation of cellular technology that continues to evolve through 5G-Advanced. 6G is still a research and standardisation effort. Research areas include AI-native operations, sensing and communications, new spectrum options, higher performance and better energy efficiency. Commercial 6G depends on standards, spectrum, devices and equipment development.
Is Starlink commercially available throughout India after IMC 2026?
Reported gateway infrastructure does not mean commercial service is authorised. As of October 10, 2026, reports said Starlink was still waiting for final regulatory and security clearances. Rely on formal launch announcements and published service terms.
What is a sovereign cloud?
A sovereign cloud gives stronger control over data location, jurisdiction, access and operational governance. Review the actual contracts, technical controls, backup locations, support access, encryption and audit arrangements instead of relying on the label.
How can AI improve telecom cybersecurity?
AI can help identify anomalies, correlate alerts and prioritise incident response. It also produces false positives and can be abused by attackers. It should complement identity controls, segmentation, secure APIs, patching, monitoring and tested incident-response procedures.
What are network APIs?
Network APIs expose selected operator capabilities through software interfaces. Examples include Quality on Demand and SIM Swap checks. They can support enterprise services and fraud-risk analysis. They do not create unlimited network capacity or replace broader security controls.
Comments