Vehicular Communication Networks in the Automated Driving Era

Shan Zhang, Jiayin Chen, Feng Lyu, Nan Cheng, Weisen Shi, Xuemin ShenIEEE Communications Magazine 2018 · DOI

Abstract

Future automated driving systems must transition from "isolated islands" of intelligence to a coordinated network. While current research focuses on on-board sensing (LIDAR, RADAR, cameras), these systems face physical limitations like line-of-sight and weather interference. This paper proposes a synergy between on-board sensing and Vehicular Communication Networks (VCNs) to enable "swarm intelligence" and cloud offloading, significantly improving safety and traffic efficiency.

The Limitations of Sensing-Centric Driving

Current automated vehicles rely heavily on local hardware for environment perception. However, sensors like LIDAR and cameras are restricted by a ~200m range and can be blinded by heavy snow or large obstacles. Moreover, the computational load of processing high-definition (HD) maps—which can reach ~5 GB/km—and real-time sensor data puts immense pressure on local vehicle resources. VCNs offer a way to break these barriers by providing a "bird's-eye view" through cooperative sensing and offloading heavy tasks to edge or cloud servers.

A Framework for VCN-Enhanced Intelligence

The paper categorizes VCN-enhanced intelligence into three distinct levels:

  1. Individual Driving Intelligence: Enhances a single vehicle's perception by sharing camera feeds from surrounding cars and streaming HD map "tiles" from the cloud.
  2. Cooperative Swarm Intelligence: Vehicles share intentions (e.g., lane changes or emergency braking) to coordinate movements proactively rather than reactively.
  3. Friendly Serving Intelligence: Integrates vehicles into the smart city ecosystem for better passenger experiences and city monitoring.
Overview of V2X communication modes including V2V, V2I, V2C, and V2D in a smart city environment.

As shown in the figure above, the V2X (Vehicle-to-Everything) ecosystem encompasses various communication modes, including Vehicle-to-Vehicle (V2V), Vehicle-to-Infrastructure (V2I), and even Vehicle-to-Drone (V2D), creating a hyper-aware environment.

Technical Requirements and Categorization

To support these intelligence levels, VCNs must meet stringent performance requirements. These are categorized by scale and urgency, ranging from micro-scale emergency alerts to macro-scale traffic density monitoring.

Categorization of driving-related contents and their corresponding V2X requirements and modes.

Safety-critical "Micro-scale" events, such as brake alarms or pedestrian detection, require ultra-low latency (≤ 100 ms) and high-frequency updates (1-10 Hz). In contrast, "Macro-scale" traffic management, such as monitoring traffic jams or parking lot availability, can tolerate slightly higher delays (≤ 500 ms).

Case Study: Distributed Intersection Coordination

To validate the VCN approach, a Python-based simulation of a four-way intersection was conducted. The study compared traditional traffic lights against a distributed V2V coordination scheme where vehicles calculate their own passing order based on real-time kinematic data.

  • Efficiency: The V2V system reduced median delays from up to 230s down to 130s under dense traffic.
  • Low-Traffic Performance: At low arrival rates, the V2V system eliminated intersection delays entirely, whereas traffic lights forced unnecessary stops.
  • Adaptability: The networked approach handled asymmetric traffic flows much more effectively than rigid signal cycles.

Challenges and Future Outlook

Despite the potential, several hurdles remain. Security is paramount, as V2X introduces new cyber-physical attack vectors. Furthermore, maintaining ultra-reliable low-latency communication (URLLC) in high-mobility environments remains a technical challenge due to rapid channel fading. The authors suggest a unified "Software-Defined Network" approach to manage handovers between satellite, cellular (LTE-V), and roadside networks (DSRC) to ensure seamless connectivity.

Summary updated

BibTeX
@article{zhang2018vehicular,
  author = {Shan Zhang and Jiayin Chen and Feng Lyu and Nan Cheng and Weisen Shi and Xuemin Shen},
  title = {{Vehicular Communication Networks in the Automated Driving Era}},
  journal = {IEEE Communications Magazine},
  year = {2018},
  doi = {10.1109/mcom.2018.1701171},
}

Loading related content...

Figure