Network Automation Critical for AV Deployment, Claims Huawei
Artificial intelligence (AI) capabilities across all levels of mobile networks will be required to enable all-scenario autonomous driving, according to Chinese telecommunications giant Huawei.
In a 30-page white paper, it laid out its vision for the mobile network capabilities required to make widespread use of autonomous vehicles a reality. It identifies seven autonomous driving scenarios from three perspectives: network operation and management lifecycle; digitalization; and scenario-specific total cost of ownership contribution.
It claimed that critical to making higher level automation on public roads widespread will be very strict requirements for low latency, down to millisecond level, and a network based AI to maintain real-time control of radio resource management. The report noted security and reliability, which it calls the most important mission of the network, requires quick alarm detection and immediate fault healing. It also analyzes five automation capabilities required in mobile networks, organized under what Huawei calls the POBSI model including programming, online, bridging, sensibility, and intelligence factors.
The seven scenarios outlined in the paper include base station deployment, feature deployment, network performance monitoring, fault analysis and handling, network performance improvement, home service provisioning, and network energy saving. Particularly critical will be reliable network performance monitoring, which requires the structured processing of reams of multi-dimensional data, including network data, engineering parameters and maps.
Huawei noted it is important to take an approach enabling automation to be implemented immediately, in order to deliver benefits for existing 4G networks as well as first-phase 5G. “The driverless car may not be fully in use until the 2030s but operators cannot wait that long for the autonomous driving network,” the report cautions.
In order for assisted driving to work, vehicles have to collect and process, while on the move, all kinds of data about the world around them, especially where other objects are at any given moment and how fast they are moving. That kind of precision and, especially, low-latency can only be achieved with the new 5G technology, claims Huawei.
LTE 4G is already in use by several manufacturers for non-safety critical V2V and V2X applications, such as exchanging, but not for time sensitive information including weather, road conditions and traffic data directly between vehicles.
Huawei further claims that only the arrival of 5G, offering latencies as low as 1 millisecond, will challenge DSRC V2X technologies favored by the auto industry’s biggest players, Toyota and Volkswagen who view this a more robust solution considering the existing major coverage problems with 4G.
— Nathan Eddy is a filmmaker and freelance journalist based in Berlin. Follow him on Twitter.
