Ericsson and SoftBank deploy machine learning radio network

SoftBank and Ericsson are upgrading radio access networks using a machine learning algorithm that makes it possible for cell overlap and signal strength to be analysed.
Written by Corinne Reichert, Contributor

Ericsson has announced using machine learning capabilities to upgrade Japanese telecommunications carrier SoftBank's radio access network in the Tokai region.

The machine intelligence-powered service reduces telco operational expenditure and improves network performance by grouping cells and analysing cell overlap, signal strength, and receive diversity to make decisions on carrier aggregation between cells, Ericsson said.

"This highly complex task was ... solved by implementing a cutting-edge design concept based on network graph machine learning algorithm (community detection) that Ericsson has now patented," the networking giant said.

"SoftBank was able to automate the process for radio access network design with Ericsson's service. Big data analytics was applied to a cluster of 2,000 radio cells, and data was analysed for the optimal configuration."

Head of Managed Services Peter Laurin said Ericsson is investing heavily in machine learning technology for the telco industry, citing strong demand across carriers for automated networks.

To support this, Ericsson is running an Artificial Intelligence Accelerator Lab in Japan and Sweden for its Network Design and Optimization teams to develop use cases.

In introducing its suite of network services for "massive Internet of Things" (IoT) applications in July last year, Ericsson had also added automated machine learning to its Network Operations Centers in a bid to improve the efficiency and bring down the cost of the management and operation of networks.

According to SoftBank's Tokai Network Technology Department radio technology section manager Ryo Manda, SoftBank is now collaborating with Ericsson on rolling out the solution to other regions.

"We applied Ericsson's service on dense urban clusters with multi-band complexity in the Tokai region," Manda said.

"The positive outcome exceeded our expectations and we are currently proceeding in other geographical areas with the same method and close cooperation with Ericsson."

The radio access network design as a service solution is able to support LTE networks.

SoftBank and Ericsson have also been working on 5G proofs of concept, kicking off a trial across the 4.5GHz spectrum band in August last year. The end-to-end trial utilised two 5G New Radios, virtual RAN (vRAN), virtual Evolved Packet Core (vEPC) technology, beamforming, Massive Multiple-Input Multiple-Output (Massive MIMO), and test support services.

SoftBank has also run 5G trials with ZTE, in June last year signing a partnership to trial 5G over sub-6GHz spectrum at 4.5GHz across Tokyo.

ZTE and SoftBank had been collaborating on pre-5G technology research and development including Massive MIMO technology and 5G New Radio.

Related Coverage

Ericsson: AI is key to fixing network complexity from 5G, IoT (TechRepublic)

New technologies will mean more complex networks, and Ericsson is using artificial intelligence to address optimizations and efficiency.

Verizon and Ericsson break gigabit record with 1.07Gbps speeds on 4G (TechRepublic)

In an industry first, the telecom providers hit over 1 Gbps on 4G LTE in a lab trial. Here's how.

Ericsson posts net loss of 700m SEK for Q1

Ericsson has reported its first-quarter results for 2018, recording a net loss of 700 million Swedish Kronor (SEK), an 18.5 percent improvement on last year's 10 billion SEK net loss, on net sales of 43.4 billion SEK.

NBN announces 5G trials with Ericsson

NBN is this week trialling 5G with Ericsson in Melbourne in an effort to explore potential upgrade paths of its fixed-wireless service.

Mobile device computing policy (Tech Pro Research)

Mobile devices offer convenience and flexibility for the modern workforce - but they also bring associated risks and support issues.

Editorial standards