Data load: mobile Internet in Russia will be accelerated with the help of AI
The use of artificial intelligence will increase the speed of the mobile Internet in Russia. Plans for its implementation by operators on cellular networks are discussed in the materials of the Ministry of Digital Affairs. Neural network algorithms are able to more efficiently redistribute Internet traffic flows, which, according to the conducted tests, allows for a 20-30% increase in network bandwidth and a halving in data transmission delays. This is extremely important for the operation of 5G networks, which are planned to be deployed in Russia in the near future, the ministry notes. How AI operators in other countries use it is described in the Izvestia article.
Why do mobile operators need artificial intelligence
Artificial intelligence is planned to be massively implemented in the telecom industry. This follows from the materials of the Ministry of Finance aimed at interdepartmental coordination within the framework of updating the Strategy for the development of the communications industry until 2035. In particular, AI should increase the speed of the mobile Internet: the draft amendments provide for the creation of national AI platforms for managing communication networks, including systems for dynamic distribution of the radio frequency spectrum and forecasting network load.
"Russian telecom operators have already implemented pilot projects for the implementation of AI network management systems, which demonstrate an increase in network capacity by 20-30% and a reduction in delays by 50%," the materials of the Ministry of Digital Affairs indicate.
In particular, it is discussed that AI can be used for dynamic control of the radio frequency spectrum, automatic optimization of base station parameters and predictive network load planning, primarily for the development of promising 5G and 6G networks, the ministry said.
However, as a source in the communications market noted, AI will allow for more efficient management of existing LTE devices. The speed of mobile Internet using AI can be increased by adjusting the load on networks: for example, redirecting Internet traffic to free communication channels and freeing overloaded ones, he explained.
At the same time, increasing the capacity of telecom networks is not the only area of application of AI in the communications industry, which is mentioned in the materials of the Ministry of Digital Affairs. Other tasks include detecting fraudulent calls, protecting against DDoS attacks, as well as managing equipment energy consumption and other functions.
Operators are already implementing neural networks into their infrastructure, the companies told Izvestia. The use of AI in network management will improve the continuity and quality of communications, increase the speed of networks, as well as speed up the processing of client requests and troubleshooting tenfold, MTS listed.
"AI primarily automates and ensures the self—organization of networks, and through automated data analysis and dynamic resource management, AI solutions make it possible to predict load, prevent incidents, and automatically fix problems," they noted.
The company has been operating an AI-powered frequency allocation system for several years now, which increases bandwidth through more efficient use of the frequency spectrum. In addition, MTS also uses AI technologies in the planning and construction of mobile networks, they added.
In all regions, T2 already uses the SON software and hardware complex. The system automatically analyzes network operation parameters and optimizes the interaction of base stations, ensuring that connections are maintained automatically. Previously, these tasks required the participation of engineers, a representative of the operator told Izvestia.
In addition, a number of subjects use machine learning methods to analyze large amounts of data in a mobile network, including assessing signal strength and quality. This makes it possible to use network resources more efficiently and redistribute subscribers by frequency during peak periods, the company noted. As a result, the data transfer rate increases and the risk of network overload decreases. According to representatives of T2, AI solutions will also allow analyzing the network status in real time, predicting deterioration in the quality of services, detecting anomalies and preventing emergency situations before they affect subscribers.
— AI allows operators to increase the speed of mobile Internet on existing networks through smarter network planning, management of radio resources, traffic and network configuration. This approach is also available in LTE networks," Beeline noted.
The company also uses SON, which allows the network to adjust itself to different conditions: load, interference, equipment failures. In 2026, the operator launched the AI Engineer network agent, which made it possible to speed up the analysis of technical events, reduce the burden on specialists and increase the stability of the infrastructure, according to the company.
— The use of AI is relevant for both 5G and 6G, as well as for LTE. We see the potential of AI in communication networks: neural networks can take over the management of complex, dynamic processes that are quite laborious for a person or simple algorithms to control in real time," Megafon noted.
How AI works on cellular networks
The main type of AI models used by mobile operators is called LSTM (Long Short-Term Memory), Leonid Konik, a partner at ComNews Research, told Izvestia. This is a type of neural network that is used to process and predict sequences of data and time series, he explained.
— The key profile of LSTM application in cellular networks is smart resource allocation. Instead of manually configuring, AI predicts the types and volumes of traffic, dynamically allocating a frequency band for different tasks: video transmission, IoT, autonomous cars, and more," the expert said.
At the beginning of 2026, Ericsson announced that the use of LSTM could increase network bandwidth by 20% and spectral efficiency by 10%, while simultaneously improving the quality of service for users with "heavy" traffic (such as video), increasing geo—positioning accuracy and coverage density, Leonid Konik continued. These results can also be achieved in LTE networks, but the main effectiveness of the LSTM model is shown in 5G networks, he added.
The use of AI will really increase the speed of mobile Internet in the Russian Federation by increasing bandwidth by 20-30%, as shown by operator tests, agrees NTI wireless expert Oleg Yablokov.
— On cellular networks, AI manages traffic distribution, optimization of radio parameters and fault diagnosis in real time, automatically reconfiguring resources. At the same time, the effect is achieved not only on 5G networks, but also on existing LTE — foreign cases, for example, from Ericsson and KDDI, record an increase in 4G bandwidth," he told Izvestia.
Russian operators are already using AI for monitoring and analytics, and in the world, Deutsche Telekom and Google Cloud have reduced the time to eliminate major incidents from hours to a minute, which proves the high effectiveness of these solutions today, Oleg Yablokov noted.
Of course, the introduction of artificial intelligence on cellular networks is important — it will really increase their bandwidth, said Denis Kuskov, CEO of TelecomDaily. But a radical increase in mobile Internet speed will occur only after the launch of 5G networks in Russia, he concluded.
Переведено сервисом «Яндекс Переводчик»
