ITU-R has defined three types of 5G application scenarios in June 2015: eMBB
(Enhanced Mobile Broadband), mMTC (Massive Machine Type Communication),
and uRLLC (Ultra Reliable & Low Latency Communication). In addition, the
capability requirements for 5G networks are defined from eight dimensions, such
as throughput, delay, connection density, and spectral efficiency improvement.
Huawei predicts that there will be 100 billion connections around the world in
2025.
Huawei Wireless X Labs research shows that Cloud VR, which is rendered by cloud,
will be the development trend of VR in the future. In local VR mode, the VR
terminal needs to be connected to the local server through cables. The user
experience is poor and the cost is high. The Cloud VR implements the wireless
function of the terminal and implements image rendering through the cloud
server, this greatly reduces terminal costs and improves user experience. Cloud VR
poses higher requirements on mobile networks, including bandwidth and latency.
For example, entry-level experience requires 100 Mbit/s bandwidth and 10 ms
latency, while ultimate experience requires 9.4Gbps and 2 ms latency, only 5G
networks can meet VR experience requirements. Currently, the application
scenarios of VR are video and Gaming, which will be expanded to more application
scenarios in the future.
Delay: The delay of the 4G network is less than 50 ms, which is equivalent to half of
the 3G network. However, applications such as automatic driving require a much
lower latency than 4G networks.
Take automatic driving as an example. Under the existing 4G network delay
conditions, car with a speed of 100 km/h requires 1.6 meters to stop down
start from discovering obstacle until startup braking system. This is the
distance between life and death.
Under the 5 G network condition, the distance between the vehicle with the
same speed can be reduced to 3.3 cm, which is expected to reach the ABS
level.
5G can achieve a ultra-low latency of 1 ms, which will increase the response
speed of 5G networks by 50 times than the 4G networks.
Currently, automatic driving mainly uses devices such as video cameras, radar
sensors, and laser rangefinder to operate.
Level 0: Manual driving, no driving assistance system, only reminder
Level 1: Assisted manual driving, a single speed or steering control
automation (such as fixed speed cruise, ACC).
Level 2: In some automatic driving scenarios, the vehicle speed and steering
control can be implemented automatically. But the driver must always keep
monitoring (for example, maintain at the middle of the lane).
Level 3: Conditional automatic driving, can be hands off, driver monitoring
system and intervene if necessary.
Level 4: Advanced automatic driving, can be eyes off. In some predefined
scenarios, drivers do not need to be involved.
Level 5: Full automatic driving, full automation, no need for driver (driverless)
4G cannot meet future applications requirements, regardless of the delay,
throughput, and the number of connections.
Based on the above example, the future services have different requirements on
the network. Therefore, the future network must be flexible.
