The use of artificial intelligence in telecommunications networks.

The increase in the number of subscribers and the volume of consumed services, including internet traffic, is forcing companies to increase the number of resources necessary to ensure high-quality services and uninterrupted operation of devices. To optimize the growing costs of maintaining and servicing equipment, cellular operators are using artificial intelligence-based telecommunications software.

In addition to chatbots and subscriber service assistants, neural networks and machine learning are used to manage telecommunications equipment for network optimization and support, cost reduction.

Preventive maintenance.

Without using AI-based tools, network monitoring was performed in several stages:

• Network engineers analyzed data from various sources.
• They identified trends, such as overload and packet loss.
• Then they performed packet loss analysis and delay cause analysis.

After completing these procedures, the causes of anomalies were eliminated.

The use of AI allows performing these stages concurrently and in real-time, which significantly speeds up the troubleshooting process.

In particular, Vodafone uses machine learning to detect and correct anomalies before they affect the quality of services provided. The anomaly detection monitors services and independently identifies cells that exhibit unusual behavior. This allows engineers to quickly detect and resolve situations such as equipment overload, interference, unexpected delays, call processing issues between cells or failures. In addition, the system identifies patterns of changes. The company expects that the solution will detect and resolve 80% of all mobile network problems, particularly in terms of throughput.

Energy efficiency.

A significant portion of telecommunications equipment consumes energy regardless of the load level, such as the number of calls or the volume of internet traffic.

In response to rising electricity costs, operators are starting to use artificial intelligence to optimize energy consumption by equipment.

Telefónica Spain became the world’s first operator to use AI-based telecommunications software for energy saving in a 5G deployment configuration platform. It is based on artificial intelligence and machine learning algorithms. Energy-saving functions are activated in the mobile network, enabling the disconnection of equipment components and infrastructure during low traffic conditions. The company achieved savings of up to 8%, reaching 26% during low traffic hours. With continuous operation throughout the day, energy consumption was reduced by up to 16%.

In addition to supporting operability and energy efficiency, AI is used for security and fraud prevention, network design and planning and infrastructure optimization.

According to the International Data Center, 63.5% of telecommunications companies are actively implementing AI with the aim of improving the quality of network operations.

Epol Soft LLC has been developing telecommunications software for over 15 years. More detailed information about projects and competencies can be found on the website in the Telecommunications section.