Autonomous driving is one of the most remarkable usages of 5G new radio (NR). Sure, 5G is here (or rather there), but autonomous driving is not anywhere — yet.

Traditional communication services for navigation, entertainment, security, vehicle maintenance and other services cannot meet the fundamental needs of high concurrency and low latency of next-generation vehicle-to-everything (V2X) operations. V2X will connect autos, road infrastructure, smart devices and even pedestrians on a single network, which seamlessly accommodates the safe and efficient transit of all connected entities.

Even though 5G deployment is sparse, networking and automotive engineers have already identified some of the minimum needs of a 5G network that will turn autonomous autos into a reliable, safe and disruptive transportation technology.

Requirements of V2X service

In the V2X services, there are five major characteristics and requirements defined: high concurrency, high real-time, high mobility, data heterogeneity and infrastructure sharing.

High concurrency: Reliable V2X service requires millions of simultaneous data transmissions, with congruency along nationwide and potentially even global networks. This ensures vehicles make decisions using the most up to date and relevant information. According to European standards, concurrency is most apparent for cooperative awareness messages (CAMs) and decentralized environmental notification messages (DENMs). In Chinese standards, V2X elements reliant on concurrency include basic safety messages (BSMs), signal phase timing messages (SPATs), navigation maps, road side information (RSI) and road safety messages (RSMs).

High real-time: The real-time requirements in the V2X services are high. According to the requirements defined in the standards, the delay of assisted driving cannot be more than 100 ms, and, in the case of autonomous driving, it cannot be higher than 20 ms.

High mobility: This means providing access and computing services in the edge. V2X needs to support continuity among all the edge service nodes, including switching between different V2X nodes.

Data heterogeneity: V2X operational data has multiple heterogeneities, which includes standard data sets as previously mentioned (BSM, SPAT), and non-standard data including camera data, light detection and ranging (lidar) data, millimeter-wave radar data and vehicle navigation data, among others.

Infrastructure sharing: The infrastructure of V2X is diverse and the construction cost is also very high. To better share the core resources, it is necessary to develop the comprehensive environment for the V2X network to support the sharing of real-time data.

Architecture of V2X platforms

The V2X platform includes applications from the upper layers to provide major functions, such as high concurrent data access, fusion analysis, high-precision positioning, open network capability, edge computing and guaranteed business continuity to meet the needs of assisted driving and autonomous driving.

With the emerging availability of 5G in many countries, and the anticipated V2X environment, some standards and industrial organizations are actively promoting an end-to-end standard system of V2X, and the standardization work to make the platform real is steadily advancing.

For example, the 5G Automotive Association has examined deployment of a cellular vehicle-to-everything (C-V2X) network architecture, as well as a V2X platform based on different application requirements. As a result, special projects are set up to carry out research on platform architecture, function and interface definition. Some other international organizations are also carrying out research on the application layer and defining the data interaction, which gradually improves platform-related standards, and accelerates the deployment and application of the V2X platform.

For example, take a sizable city with about three million vehicles (for reference, Chicago has roughly 3.5 million). The V2X platforms needs to process tens of millions of data transmissions per second, and the processing delay of a single transaction is required to be between 20 ms and 100 ms, which is challenging for data access, data calculation, data storage, data publishing and data security. The traditional architecture of "core — platform — endpoints" cannot meet the requirements of V2X. Therefore, a multi-level architecture of the V2X platform is necessary.

The V2X core platform mainly supports the network operations and provides global management functions within the network, including the operation management of the whole network, global traffic environment learning with optimization, multi-level computing function scheduling, multi-level dynamic deployment of applications and cross-regional operation data management.

The V2X regional platform mainly supports regional operations, including regional operation management, regional traffic management, regional traffic learning and optimization, regional data analysis and publishing, edge collaborative computing scheduling, V2X edge node management, and some other functions. All of these services can serve operation scenarios with high sensitivity to latency.

Finally, the V2X edge node supports V2X operations with high real-time and high bandwidth requirements within the edge range. The functions of data consolidation and sensing with dynamic map construction can serve the applications of advanced assisted driving and autonomous driving.


In order to better provide more comprehensive and accurate real-time traffic information for the V2X applications in the network, it is widely suggested that a V2X core platform serving this special network be built, so as to realize the interconnection and data sharing among government transportation platforms, enterprise application platforms and social service platforms, effectively breaking the data barrier between the platforms of the separate V2X networks.

On the other hand, to better promote the large-scale application of V2X and the interconnection between various ends and platforms, government, business and consumers need to promote the development of standards in areas such as data interfacing and operation interaction processing among endpoints, application services and the V2X platform.

To contact the author of this article, email