Successfully achieved an approximately 24% improvement in 5G downlink throughput in a demonstration at Expo 2025 Osaka, Kansai, Japan

SoftBank  and Ericsson Japan  announced that they have deployed an AI-powered, externally controlled coverage optimization system for Massive MIMO at multiple large-scale event venues in Japan, including major arenas and dome-type facilities in the Tokyo metropolitan area, which automatically enhances coverage patterns in response to traffic changes. This follows a successful commercial network trial at the Expo 2025 Osaka, Kansai, Japan .

At large-scale event venues and theme parks, user distribution varies significantly between normal periods and during events, requiring flexible base station control in response to these distributions and changes. In particular, for events subject to weather-related cancellations or schedule changes, such as fireworks displays, or where specific area access restrictions occur, conventional methods that preset coverage pattern changes by time slot were sometimes unable to respond appropriately. Therefore, a system capable of automatically controlling and optimizing coverage patterns according to the situation was needed. To address these challenges, SoftBank and Ericsson have repeatedly conducted trials to validate automated traffic optimization using AI-driven external base station control.

When deployed at base stations installed in outdoor areas of the Expo 2025, the system delivered clear improvements in communication quality, including an approximately 24 percent increase in downlink throughput, from 76.9 Mbps to 95.5 Mbps, during periods of rapid traffic fluctuation.

Building on these results, SoftBank and Ericsson have begun operating the system at large-scale event venues where user distribution varies significantly by time of day.

Overview of the System

• An external control device (server) uses user distribution data collected from base stations at one-minute intervals to automatically determine event occurrence using AI
• Dynamically and automatically optimizes the horizontal and vertical coverage patterns of Massive MIMO base stations

Overview of demonstration at Expo 2025:

(1) Location
Massive MIMO-compatible base station installation areas within the Expo 2025 venue

(2) Content
An AI model was constructed using performance results obtained when multiple coverage patterns were changed in advance as training data. Based on user distribution-related data such as Massive MIMO beam estimation information acquired by an external control device from base stations at one-minute intervals, the system automatically determined event occurrence status and switched base station coverage patterns to optimal configurations.

(3) Key Results
Downlink throughput for 5G users across the entire area during event periods improved by approximately 24 percent. This result contributed to avoiding “packet stalling” phenomena where data transmission and reception stagnate due to congestion that tends to occur during rapid traffic fluctuations.

Roles of each Company

• SoftBank
  Use case examination and demonstration evaluation of the system
• Ericsson
  Provision of the system

The two companies will continue to actively expand deployment of the system.