SoftBank, Ericsson bring machine learning to mobile network design

SoftBank, Ericsson bring machine learning to mobile network design

Japanese technology giant SoftBank has partnered with Swedish telecoms provider Ericsson to trial new radio technology that uses machine learning to design networks and improve performance.

SoftBank, whose investments in emerging technologies include Uber and robotics specialist Boston Dynamics, is a leading mobile network operator in Japan and one of the country’s largest organisations. The company has been implementing Ericsson’s machine intelligence technology for advanced radio network design in the Tokai region of the country.

Machine learning in the telecoms industry

Ericsson’s system relies on a thorough analysis of the radio network environment. This includes assessing cell coverage overlap, signal strength, and receiver diversity.

This analysis requires both heavy computational power and advanced machine learning techniques. It uses a design concept based on a network graph machine learning algorithm – a method that has since been patented.

The method groups cells – sections of land where the network is distributed – into clusters, takes statistical data from overlapping cells, and takes into account the potential to use carrier aggregation between them.

According to SoftBank and Ericsson, the work in Tokai was a “highly complex task” and a “tremendous challenge”, but the result is a design that cuts operational expenditure and leads to better network performance.

Compared to traditional network design methods, it cut lead time by 40 percent.

Ryo Manda, radio technology section manager at SoftBank’s Tokai Network Technology Department explained, “We applied Ericsson’s service to dense urban clusters with multi-band complexity in the Tokai region. The positive outcome exceeded our expectations, and we are currently proceeding in other geographical areas with the same method and close cooperation with Ericsson.”

Automating the network design process

SoftBank was able to automate the process for radio access network design using Ericsson’s service, which applied big data analytics to more than 2,000 radio cells to come up with the best possible network configuration.

Peter Laurin, head of managed services at Ericsson, believes the Tokai trial represents the future of network design. “There is huge potential for machine learning in the telecoms industry and we have made significant investments in this technology,” he said.

“It is very exciting to see that the new methods have been successfully applied in SoftBank’s network. There is a strong demand for this type of solution, and deployments of this service to other tier-one operators in other regions are ongoing.”

Internet of Business says

Ericsson’s radio network design method supports LTE networks and is currently being sold as a service. Ericsson provides services in networks, IT and network design, and optimisation. The network design for SoftBank is part of the Swedish company’s wider strategy to take machine learning into new industries.