Research on Laser Counterfeit Command Attack Mechanisms for In-vehicle Voice Interfaces in Intelligent Connected Vehicles

Abstract

With the evolution of human-machine interaction in intelligent connected vehicle (ICV), voice control has become a core feature. However, the associated security risks are increasingly evident, particularly remote, covert audio injection attacks such as laser voice attacks, which can bypass traditional access controls and directly inject malicious commands into onboard microphones, posing a serious threat to vehicle safety. This paper first conducts an in-depth analysis of the attack surface of in-vehicle voice control systems, focusing on the technical principles, attack chain, and specific security risks posed by laser voice attacks to vehicles. It also outlines the existing defensive measures available in the market, analyzes the advantages and limitations of current technologies, and proposes a solution based on the spectral consistency of voice signal physical layers and environmental features to enhance the overall security of the system.