Ericsson has landed a deal from NTT DoCoMo that will see it provide AI-based radio access network (RAN) optimization tools to the Japan-based operator.

NTT DoCoMo is to use the Swedish vendor's AI-based Performance Diagnostic tool to classify cell performance issues and recommend changes. Ericsson said trials have already been carried out in 12 cities in Japan, and that the commercial operation to cover the nationwide network is planned for March 2020. The vendor said the tool makes use of machine learning and complex problem-solving techniques to identify and resolve issues.

Ericsson added that the Performance Diagnostic tool is one fruit of its longstanding collaboration with NTT DoCoMo that has recently been focusing on 5G proof-of-concept activities. NTT DoCoMo has yet to reveal its 5G timetable, but already faces aggressive challenges from rivals Rakuten Mobile and SoftBank.

AI-based tools are regarded as key to real-time diagnostics, and proactive network health strategies, in 5G. Indeed, network management is seen as one the first areas where AI will be used in telecom. For example, BT said last year that AI and machine learning are already key components in the closed-loop control of its programmable network and are used to detect anomalies and take pre-emptive action.

As stressed by Peter Laurin, executive vice president and head of business area managed service at Ericsson, the vendor's collaboration with NTT DoCoMo "highlights the real benefits and value that AI-based and machine leaning solutions can bring to communications service providers' network operations."

For more on this topic, see:

— Anne Morris, Contributing Editor, Light Reading