5G, millimeter wave radar and UWB accelerate the deployment of autonomous driving

The so-called autonomous driving technology basically collects all necessary information through sensors installed on the car body. Therefore, relying solely on these sensors may meet the requirements of Level 3, but when it is expected to achieve Level 5 fully autonomous driving , more information is needed. For example, when driving, the driver cannot see the dangers and surrounding road conditions. The information obtained by sensors alone is far from enough.

At this stage, autonomous driving technology has reached Level 3 and can travel on highways. Although various advanced automatic technologies are being actively developed, the successful introduction of autonomous driving requires the support of three advanced wireless technologies: 5G , millimeter wave radar and UWB . Upgrading autonomous driving to Level 5

through 5G

Compared with traditional 4G, 5G has three main features: high speed and high capacity, ultra-low latency and multi-point simultaneous connection. For example, in terms of the peak rate of communication speed, 4G is 1Gbps, but 5G has been greatly improved to 20Gbps, so it can meet the transmission of 4K/8K or other high-resolution images. The delay speed will also be greatly reduced to 1 millisecond, which is 1/10 of 4G. The number of simultaneous connections has also been greatly increased to 1 million units/k㎡, which is ten times that of traditional systems. Therefore, through 5G communication technology, it can help to quickly improve the autonomous driving capability to Level 4 or even Level 5.

At present, the global autonomous driving technology-related industry has been working hard to develop a vehicle-to-vehicle communication system using 5G communication. For example, Japan's SoftBank developed the world's first high-reliability, low-latency vehicle-to-vehicle communication system required for automatic safety driving before international standardization, and successfully applied it to trucks to drive in formation with subsequent vehicles on the highway.

This new generation of safe driving system connects three trucks driving at high speed through 5G communication technology and communicates directly between wireless terminals. Because direct communication between terminals does not require a base station, the delay in signal transmission is only 1/1000 second, allowing low-latency transmission for communication between vehicles.

At the same time, SoftBank Japan also uses the low-latency broadband transmission characteristics of 5G to transmit images from the rear truck camera to the vehicle in front, so that all vehicles using 5G CAV communication can observe the situation around the rear vehicle in real time.

The platform called "extended sensors" by SoftBank Japan can share information required to maintain distance and automatic tracking between vehicles. After actual driving on the new Tokyo-Nagoya Expressway, it has been proven that this technology is trustworthy for safe driving. Therefore, it is also considered by the Japanese industry to be very useful for the standardization of "Connected Autonomous Vehicle (CAV)", and it is also a new technical concept necessary to promote the advancement of autonomous driving technology.

Use V2X to connect cars to various external things

In order to realize the Internet of Vehicles, there has been more and more discussion about " V2X (Vehicle to Everything)" recently. With the emergence of 5G, people's expectations for 5G's application in V2X are getting higher and higher (Figure 1).

Figure 1: People expect that 5G will be able to meet the requirements of V2X communication network technology. (Source: Japanese Ministry of Internal Affairs and Communications White Paper; compiled by the author)

An obvious example is 3D map data. To achieve autonomous driving, more detailed data than conventional car navigation systems is required. However, it is not practical to store such a large amount of information in the car, so frequent updates of 3D map data through communication are essential, which requires high-speed, high-capacity communication between the vehicle and the network (V2N). In addition, in autonomous driving above Level 3, the driver is not the driver, but the system.

To ensure safe driving in this situation, remote monitoring and management, or even remote operation, is required. In the past, the biggest problem was the time lag or delay in communication. However, using the ultra-low latency 5G network, these delay issues can be greatly reduced.

5G's multi-connectivity also provides various intelligent capabilities for autonomous driving, such as vehicle-to-vehicle connection (V2V), which enables automatic platoon driving and advanced accident prevention. And connecting to roadside infrastructure such as traffic lights (V2I) can achieve safe and efficient driving. In addition, if the connection with pedestrians' smartphones and other devices (V2P) is connected, the vehicle can be controlled to automatically dodge or stop when pedestrians suddenly enter the dangerous area of ​​the vehicle body.

On the other hand, the application of 5G networks still requires advanced technology. Currently, 5G uses two frequency bands, Sub 6 and millimeter wave, for communication. Traditional 4G is below 3.6GHz, Sub 6 is at 3.7 GHz and 4.5GHz, and the frequency of millimeter wave is higher, reaching 28GHz. Therefore, the development of 5G standard communication equipment requires advanced technologies such as millimeter wave, antenna design and heat dissipation (Figure 2).

Figure 2: 5G uses the Sub 6 (3.7 GHz and 4.5 GHz) and millimeter wave (28 GHz) frequency bands for communication. (Source: Japanese Ministry of Internal Affairs and Communications White Paper; compiled by the author)

In addition, there are some special technical challenges in the application of V2X, because the requirements of V2X are different from traditional mobile communications, requiring very high reliability and extremely low latency, and the network configuration is more complex, as well as the connection and communication between terminals. Therefore, the development of V2X requires not only advanced wireless technology, but also deep mobile communication expertise (Figure 3).

Figure 3: V2X requires advanced mobile wireless technologies as a communication basis. (source: Fujitsu; author’s compilation)

Development of connected car networks requires careful performance evaluation and simulation

Due to the multiple superposition of various factors, unexpected problems often occur in the development of vehicle-connected networks, which have never been experienced before. For example, when multiple vehicles are driving on the road, radio interference or network performance degradation occurs. In the real environment, vehicles are objects that move at high speeds, and these large amounts of traffic are always moving in staggered directions, so the communication environment and conditions are changing all the time. In order to accurately grasp these changes and related conditions, it is necessary to construct a simulation model (Simulation Model) and an actual field environment (Field) for testing.

By using a simulation model that complies with mobile communication and V2X standards, the number, location and direction of vehicles can be preset, and the various performances of the vehicle-connected network can be further verified. And after collecting data from field test results, the accuracy is further improved, development is accelerated and the verification cycle is shortened.

In addition, when conducting radio testing, it is necessary to certify according to relevant laws and regulations, which is very cumbersome and time-consuming in the process. Therefore, how to cooperate with excellent wireless technology and mobile communication partners is also the key to the smooth development of connected cars.

When verifying a wireless network Telematics Control Unit (TCU), compatibility with the various sensors and Electronic Control Units (ECU) already installed in the vehicle must be considered. Each sensor or ECU has different requirements for network conditions, such as latency, communication cycle, and communication capacity. For example, in order to find the best network requirements for each situation, whether traditional 4G is sufficient or 5G must be used, a prior integrity assessment is required. In addition, network slicing and MEC (mobile edge computing) technologies are also very important for establishing the most appropriate and flexible network for each application object unit.

More effort and time are required in the early stages of the development of the connected vehicle network .

In addition, for 5G and millimeter wave radars in V2X, the millimeter waves used have extremely short wavelengths and high transmission losses, so antenna design, circuit mode loss and noise have a great impact on performance.

The millimeter waves used in the 5G architecture of V2X have extremely short wavelengths and high transmission losses, so the design of antenna design, circuit mode loss and noise has a great impact on the communication capabilities of the vehicle network. On the other hand, the short wavelength also means that the size of the circuit is very small, and each side of the printed circuit board is only a few millimeters wide. The patten width must also be accurate to the micron level, which requires advanced packaging technology. It is difficult to manually adjust such high-resolution millimeter wave circuits during the prototype design time, which also means that additional efforts and time are required during the design and prototype development period.

In order to solve these problems, how to effectively use technologies such as electromagnetic simulation becomes very critical, because constructing a high-precision simulation model is very important for the accurate and effective design of millimeter wave components.

However, the establishment of a simulation model cannot be achieved overnight. Only by gaining experience through repeated verification in millimeter wave development can the accuracy of the entire network platform be improved.

Intelligent cockpit instrumentation for autonomous driving