5G-Advanced Promises Step Change for Extended Reality

At the same time, the network layer senses service information and then protects the integrity of service frames, while implementing differentiated scheduling of different data flows, with a view to providing an end-to-end frame-level quality of service (QoS) guarantee.

Using these techniques, the cross-layer collaboration technology can preferentially guarantee high-priority image frames. With an XR service, the first frame in a group of pictures (known as the I frame) contains the complete image and is the most important frame.

The following P frames (predicted frames) are encoded and decoded based on the I frame. By sensing the identification information at the service layer, a 5G-Advanced network can prioritise the I frames, discarding less important data when the network is congested to ensure the I frames gets through (see graphic).

During data transmission, each video frame is divided into multiple IP data packets. If one or more of these IP data packets are lost, decoding of the corresponding video frame will fail. By sensing frame integrity and frame-level QoS information, an Advanced-5G network could schedule data in a single frame at the same time or discard packets if necessary. For example, if an IP data packet is lost, other IP data packets belonging to the same video frame could be discarded, so as to avoid transmitting useless data, thereby saving network resources.

“Through these methods, the network can provide end-to-end frame-level QoS guarantee more effectively,” says John Gao.

Another key connectivity challenge for XR services is the fact that they depend on multiple types of information flow data, such as audio, video, control signals, and information collection signals. The service provider needs to coordinate the transmission of these multiple types of service flows to ensure that the time difference between the service flows and the terminal side is within the acceptable range.

The cross-layer collaboration technology is designed to support synchronous transmission of multiple streams. “By sensing data flow information at the XR application layer, the Advanced-5G network maps data flows with different QoS requirements to different QoS flows by using hierarchical QoS mapping, and combine differentiated scheduling, multi-QoS flow association and collaboration, and multi-QoS flow synchronisation enhancement,” explains John Gao. “The eNodeB (the base station) performs more refined scheduling and transmission for XR services to further improve the user experience of XR services.”

With a conventional 5G network, the minimum granularity for QoS assurance is QoS flows, according to Huawei, which means packets with different priorities in the same QoS flow cannot be identified
and differentiated.

Huawei and China Mobile plan to test how well 5G-Advanced networks will support XR services in outdoor scenarios, such as inside vehicles, where cellular networks are the only realistic connectivity option. “Fixed broadband and Wi-Fi cannot be used in a wide range of geographical and mobile scenarios,” notes John Gao. “Therefore, 5G-Advanced will be the inevitable choice for XR application explosion.”

China Mobile and Huawei will continue to work with industry partners, such as chip providers, device vendors and content developers, in a number of ways. They plan, for example, to further refine the network requirements of different Metaverse service scenarios, while exploring how 5G-Advanced technologies can support XR services in the consumer and business markets, underpinned by sustainable business models for each player in the value chain.

“As we know, 5G drives the development of video services towards HD and immersive. From HD/4K, cloud gaming, VR/AR, to more immersive metaverse requirements, new business forms are constantly developing, which brings new requirements to network and computing infrastructure,” concludes Huang Yuhong, General Manager of China Mobile Research Institute.

“The requirements of ultra-low latency, ultra-large bandwidth, super virtual-real-time synchronization, and ultra-real- time interaction in the Metaverse bring unprecedented challenges and are also an important driving force for infrastructure development.”