SoftBank has achieved a major advance in 5G radio access network (RAN) performance by using quantum computing, specifically an Ising machine, to optimize network configurations in a Tokyo proof-of-concept. The core of their approach targets carrier aggregation (CA), a technology that boosts data speeds and capacity by letting devices simultaneously connect to multiple frequencies and base stations. As networks become denser, the task of determining the optimal CA links among base stations explodes in complexity—ten base stations yield 45 possible pairs, but since each pair can reside in a linked or unlinked state, the total possible configurations reach a staggering 35 trillion.
Conventional (classical) computing struggles with such “combinatorial optimization” in real time. An Ising machine—a special-purpose quantum system—excelled at mapping network areas into fine-grained “meshes” and systematically computing the best CA link combinations. SoftBank’s experiment divided a section of urban Tokyo into small meshes, then used quantum computation to maximize the number of meshes where CA could be enabled.
The results were significant:
Average downlink data speeds in the test area improved by about 10%.
Overall data transmission capacity increased by up to 50%.
Both the proportion of connections using CA and the volume of data handled via secondary, parallel channels rose substantially.
These gains translate directly into a better user experience: more consistent high-speed connectivity for demanding applications like high-resolution video and online gaming, and more efficient use of limited radio spectrum—crucial as traffic surges with wider 5G adoption.
Beyond CA optimization, SoftBank is planning broader applications for quantum computing across its network architecture and operations. This includes integrating with its “AI-RAN” (Artificial Intelligence–Radio Access Network) initiative, which unifies AI and network management on advanced infrastructure, and an ongoing partnership with quantum companies like Quantinuum for next-generation network and data center use cases. While a permanent, city-wide deployment is yet to be announced, the Tokyo trial proves quantum approaches can tackle previously intractable network optimization problems and offer clear performance and efficiency advantages over traditional methods.
CT Bureau