MUMBAI, Oct 11: Artificial intelligence could help telecommunications operators improve the performance of their existing fifth-generation mobile networks by 25–30 per cent without necessarily investing in additional hardware, according to Swedish telecom equipment manufacturer Ericsson.
The development could offer a significant opportunity for telecom operators in India, including Reliance Jio and Bharti Airtel, which are continuing to expand and optimise their 5G infrastructure while preparing for the next generation of mobile communications.
As demand for mobile data, cloud services, video streaming and AI-powered applications continues to grow, telecom companies are under pressure to deliver faster and more reliable connectivity while controlling operational expenditure. AI-based network optimisation could help address these competing requirements by enabling operators to manage network resources more efficiently.
Ericsson’s assessment points towards a shift in how telecom infrastructure may be upgraded. Rather than depending exclusively on new equipment to improve performance, operators could use software, data analysis and intelligent automation to extract more capacity from infrastructure that is already deployed.
The approach could become increasingly relevant as mobile networks grow more complex and customer expectations rise. However, the actual benefits would depend on network conditions, the quality of available data, the AI tools deployed and the effectiveness of their integration with existing systems.
AI offers a new approach to network optimisation
Traditional telecom network management involves monitoring traffic, identifying congestion, adjusting network parameters and allocating resources to maintain service quality. These activities require operators to analyse large volumes of information generated by network equipment and connected devices.
Artificial intelligence can assist by identifying patterns in network behaviour and recommending or carrying out adjustments based on changing demand.
For example, traffic levels can vary significantly according to location, time of day, local events and the number of people using data-intensive applications. A network that experiences heavy demand in a commercial district during working hours may face a different pattern of usage in residential areas during the evening.
AI systems can analyse these variations and help operators allocate capacity more effectively. They can also identify recurring performance problems and support faster responses to congestion or changes in traffic.
Such capabilities could improve the utilisation of existing infrastructure, reducing the need to expand physical capacity every time usage increases.
Ericsson’s estimate that AI could improve 5G performance by 25–30 per cent highlights the potential scale of these efficiency gains. The figure should not, however, be interpreted as a guaranteed improvement for every operator or every network. Performance depends on the starting condition of the network, the type of optimisation applied and the metrics used to measure results.
The distinction is important because a network’s performance includes several factors, such as download speeds, upload speeds, latency, coverage, reliability and the ability to handle simultaneous connections. An improvement in one area may not automatically translate into an equivalent improvement across all these measures.
Why the development matters for Indian telecom operators
India has become a major market for mobile connectivity, with telecom operators investing heavily in expanding coverage and improving service quality.
The growing use of digital payments, online education, video services, connected devices and cloud-based business applications has increased the importance of reliable mobile networks. Enterprises also require dependable connectivity to support digital operations, customer services and remote working.
For operators, the challenge is to meet this demand without allowing infrastructure and energy costs to rise disproportionately.
AI-based optimisation could provide a way to improve network efficiency before committing to additional hardware investments. If operators can increase the capacity available from existing infrastructure, they may be able to delay some upgrades or direct investment towards areas where demand is strongest.
The potential savings could be relevant in a highly competitive market where operators must balance network expansion with affordable services.
However, introducing AI tools also involves costs. Operators may need to invest in software, computing resources, data management, staff training and cybersecurity. Integration with existing equipment and operational systems may also require technical changes.
Consequently, the financial benefits would depend on whether improvements in performance and operating efficiency outweigh the cost of deployment and maintenance.
For large operators, the opportunity may lie in applying AI across extensive networks, where even relatively small efficiency gains can have a substantial cumulative effect.
Managing traffic as data consumption rises
One of the principal challenges facing modern telecom networks is the uneven distribution of data demand.
A network may have sufficient overall capacity but still experience congestion in specific locations. Stadiums, transport hubs, business districts and densely populated residential areas can place considerable pressure on local infrastructure.
AI systems could help operators anticipate these patterns by examining historical usage, current traffic and other network information. Such analysis may allow resources to be adjusted before congestion becomes severe.
Intelligent optimisation could also support the identification of inefficient configurations and help engineers prioritise maintenance or capacity improvements.
Another potential application is energy management. Network equipment consumes electricity even when traffic levels are relatively low. AI-based systems may help operators adjust selected network resources according to demand while maintaining required service levels.
Any such adjustment must be carefully managed. Reducing power consumption cannot come at the expense of coverage, reliability or emergency communications.
The challenge is to ensure that efficiency measures improve the overall operation of the network rather than simply shifting problems from one location or period to another.
AI and the transition towards 6G
The development also comes as the telecommunications industry prepares for the eventual transition from 5G to sixth-generation mobile networks.
Although 6G remains a developing area, research and industry planning are increasingly focused on networks that can use AI more deeply in their design, management and operation.
The next generation of connectivity is expected to involve greater integration between communication systems, computing resources, sensors and intelligent applications. Managing such infrastructure could require more advanced automation than conventional networks.
For operators, improving AI capabilities within current 5G systems may provide practical experience that can be applied to future technologies.
Rather than treating AI as an additional tool used only by network engineers, telecom companies are increasingly examining how intelligent systems can become part of routine network operations.
This transition will require reliable data, strong monitoring systems and safeguards to ensure that automated decisions do not disrupt essential services.
It will also require cooperation between telecom operators, equipment manufacturers, software developers and regulators as technical standards evolve.
Reliability and cybersecurity remain important
Despite its potential benefits, AI-driven network management introduces risks that operators will need to address.
Automated systems depend on the quality of the information they receive. Incomplete, inaccurate or misleading data can lead to poor decisions. An optimisation tool that responds incorrectly to a sudden change in traffic could potentially worsen congestion rather than resolve it.
Telecom networks are also critical infrastructure. Disruptions can affect businesses, public services, digital payments and emergency communications.
Operators will therefore need to maintain strong oversight of automated systems, establish clear procedures for responding to errors and ensure that network engineers can intervene when necessary.
Cybersecurity is another consideration. As more network functions become software-driven and interconnected, operators must protect management systems from unauthorised access and malicious activity.
Testing, continuous monitoring and clear accountability will be essential to ensure that AI improves network performance without creating new vulnerabilities.
Benefits will depend on implementation
Ericsson’s estimate offers an indication of how AI could help telecom operators make better use of existing 5G infrastructure. For India, the approach could be particularly relevant as demand for high-speed connectivity grows and companies seek to balance service quality with investment requirements.
The technology’s practical value will depend on how effectively it is deployed across different network environments. Operators will need to measure improvements against clear benchmarks, assess operating costs and ensure that service quality remains consistent.
AI is unlikely to eliminate the need for new infrastructure altogether. Areas with insufficient coverage or rapidly increasing demand may still require additional sites, equipment and spectrum resources.
Instead, intelligent optimisation could complement physical expansion by helping operators identify where investment is most necessary and where existing capacity can be used more effectively.
As the industry moves towards more automated networks, the ability to combine AI-driven management with reliable infrastructure could become an important source of operational efficiency.
For Indian telecom companies, the immediate opportunity lies in improving the performance of current networks while developing the technical capabilities needed for the next phase of mobile connectivity.