Performance Analysis of C-V2I-based Automotive Collision Avoidance

TL;DR

This paper evaluates the effectiveness of C-V2I networks for automotive collision avoidance at intersections, analyzing how server placement and vehicle automation influence safety and system reliability.

Contribution

It provides a simulation-based analysis of collision avoidance algorithms in C-V2I infrastructure, considering server location and vehicle automation levels.

Findings

Collision avoidance algorithms are effective in real-world scenarios.

Server location impacts system reliability.

Automated vehicles improve reaction times and safety.

Abstract

One of the key applications envisioned for C-V2I (Cellular Vehicle-to-Infrastructure) networks pertains to safety on the road. Thanks to the exchange of Cooperative Awareness Messages (CAMs), vehicles and other road users (e.g., pedestrians) can advertise their position, heading and speed and sophisticated algorithms can detect potentially dangerous situations leading to a crash. In this paper, we focus on the safety application for automotive collision avoidance at intersections, and study the effectiveness of its deployment in a C-V2I-based infrastructure. In our study, we also account for the location of the server running the application as a factor in the system design. Our simulation-based results, derived in real-world scenarios, provide indication on the reliability of algorithms for car-to-car and car-to-pedestrian collision avoidance, both when a human driver is considered and when automated vehicles (with faster reaction times) populate the streets.