Performance analysis of Vehicle-to-Vehicle communications for critical tasks in autonomous driving

Parra, I.; Corrales, H.; Hernandez, N.; Vigre, S.; Llorca, D.F.; Sotelo, M. A. · 2019 · OpenAlex-citations

DOI: 10.1109/itsc.2019.8917302

archive: archived pipeline: cataloged verified

Get this paper ↗ (DOI — opens at the source; we link to it, we don't host it)

Summary

This paper evaluates the performance of ITS-G5A Vehicle-to-Vehicle (V2V) communications in critical scenarios for autonomous driving, comparing them against traditional onboard sensors like cameras and radars. The research is motivated by the limitations of standalone sensor-based approaches, which often suffer from short detection ranges, high computational costs, and inability to detect vehicles beyond line-of-sight or occlusions. Conversely, while V2V offers richer information and longer range, it faces challenges regarding latency, packet loss, and reliability. The study aims to determine if V2V communications can robustly support cooperative driving tasks where sensors fail, specifically focusing on T-intersection management and elevation changes on two-way roads. The experimental design involved two vehicles: a modified Citroën C4 equipped with LiDAR, cameras, and radars, and a commercial Toyota Prius. Both vehicles were outfitted with ITS-G5 compliant V2V communication modules operating at 5.9 GHz, broadcasting Cooperative Awareness Messages at 10 Hz. The researchers conducted four experiments in real-world environments to measure detection ranges and communication quality. Metrics included Packet Delivery Ratio (PDR) and the Complementary Cumulative Distribution Function (CCDF) of Update Delay. The scenarios tested included a baseline straight road, an uncontrolled T-intersection in an inter-urban environment, an urban T-intersection, and a high-slope road with non-line-of-sight conditions. The results demonstrate that V2V communications significantly outperform onboard sensors in detection range. In baseline conditions, V2V achieved detection distances exceeding 360–400 meters, whereas cameras and radars were limited to approximately 120 and 140 meters, respectively. In T-intersection scenarios, V2V maintained reliable detection up to the limits of the test tracks (up to ~468 meters), while short-range radars were capped at roughly 50 meters. Communication reliability was high, with PDR remaining near 100% as long as line-of-sight was maintained. Packet losses and increased update delays primarily occurred when line-of-sight was obstructed by large obstacles like roundabouts or heavy vehicles. In non-line-of-sight scenarios involving elevation changes, V2V still provided superior awareness compared to sensors that require direct line-of-sight, though performance degraded due to signal attenuation. The study concludes that ITS-G5 V2V communications offer a robust solution for cooperative autonomous driving, particularly in scenarios where traditional sensors are insufficient due to range limitations or occlusions. The findings suggest that V2V can effectively complement onboard sensors, providing critical information about vehicle state and trajectory at distances far beyond sensor capabilities. However, the reliability of V2V is contingent on maintaining line-of-sight, as significant obstacles can cause packet loss. This highlights the importance of integrating V2V with sensor data to ensure robust perception in complex driving environments.

Provenance

The full processing record for this entry. Every stage of this paper's journey through the pipeline is logged — what ran, with which tool and model, how many attempts it took, and when it last completed.

Summary generated by qwen3.6-27b-prismaquant on 2026-06-26; verification: verified.

Topics

Ranked by relevance to this paper. Hover a topic for its definition.

• v2x connected vehicle