LIVE FROM HUAWEI GLOBAL MBB FORUM 2018, LONDON: David Wang, executive director of the board at Huawei (pictured), detailed his vision of a fully automated mobile network, which he explained will be needed to manage the increasingly complex architecture brought about by the addition of 5G.

In a keynote, Wang said operators already had to support 2G, 3G and 4G services across several bands and technologies, and building the 5G layer would mean the network would be “so complex.”

He explained operators must provision for “four different technologies, more than 15 different kinds of bands, a lot of different kinds of sites and a lot of different kinds of services. So it is more and more difficult to optimise and improve the network. It is more and more difficult to provide the best user experience.”

Wang explained continuing to use previous network management methods would “consume a lot of effort and the cost to improve and upgrade will get higher and higher.” He also noted there would be a large increase in power consumption.

“We have to figure out a new solution and a new way to deploy and manage the mobile network,” Wang added. “We need to adopt [artificial intelligence] AI to the mobile network.”

On that note, the executive pointed out that autonomous driving mobile networks are part of Huawei’s overall architecture for autonomous driving telecom networks – a technology it calls SoftCOM AI. According to Wang, intelligent mobile network capabilities must be built at the site, network, and cloud levels, due to the unique characteristics of mobile networks.

He also urged collaboration across the industry to define standards and promote what is described as “autonomous driving mobile network” technology.

During the session, Wang also noted Huawei had spent more than two years working on its all-IP strategy and is on target to build a fully autonomous mobile network.

Earlier this year, the company unveiled a full portfolio centred on enabling AI technology including: chips; chip enablement; a training and inference framework; and application enablement solutions.