NTT Docomo and Samsung Use AI to Predict and Prevent Mobile Speed Drops

NTT Docomo and Samsung Electronics announced on August 10 that they have jointly developed technology that uses AI to predict smartphone communication speed drops in advance and automatically switch to a better network. In a field trial conducted in Japan in January, the technology reduced the frequency of speed drops from 13.1% to 7.2%.

How the technology works

The new system analyzes anonymously collected past communication quality data from devices, known as MDT, together with user movement patterns and the services being used. The AI detects early signs of throughput degradation, which can cause buffering or quality drops during video streaming, and switches the connection to a better network before the slowdown actually occurs.

Conventional approaches typically switch networks only after quality has already deteriorated. The new technology aims to prevent interruptions during activities such as watching video while on the move or joining online meetings.

Field trial results

In experiments conducted over local 5G and commercial LTE environments in Japan, vehicles equipped with commercially available smartphones drove for about 10 minutes while simulating video streaming traffic. The results showed the frequency of speed drops was reduced from 13.1% to 7.2%.

The companies also explored a more efficient way to collect the data needed for AI optimization. Instead of uniformly gathering large volumes of wireless environment information, the system selectively collects the data needed for a given task, reducing load on the network.

Toward 6G and AI-Centric networks

Both companies are researching the technology as a core component of AI-Centric networks, in which AI autonomously operates network infrastructure, targeting the 6G era expected in the 2030s. They have jointly proposed the speed prediction and efficient data collection techniques to 3GPP, the international standardization body for mobile communications, and plan to continue standardization work toward networks that AI optimizes autonomously.

Why this matters for Japan’s mobile market

Japan’s mobile operators are preparing for a new generation of network technology while managing rising data traffic and increasingly complex spectrum environments. The collaboration between a Japanese carrier and a global device maker illustrates how AI is moving from user-facing features into the core of network operations, a shift that industry players expect to accelerate as 6G standardization progresses.

For smartphone users, the practical benefit would be fewer interruptions during video streaming, online meetings and other data-heavy activities while on the move, without requiring any changes to how they use their devices.

Source: K-Tai Watch (Impress)

Cover image generated by AI for illustrative purposes only.