Table 2 The comparison of the different types of attacks and some analysis based on various models.

From: A deep learning based intrusion detection system for CAN vehicle based on combination of triple attention mechanism and GGO algorithm

Attack type

Description

Analysis

Comparison with other works

DoS

Overwhelms a server or service with illegitimate requests, disrupting availability

Targets infrastructure or application layers, often using botnets

Similar to other DoS attacks, but specific to vehicular networks

Fuzzy

Injects random CAN ID and data values, causing confusion and potential system failure

Uses fuzzy logic to handle uncertain values, making detection difficult

Similar to other fuzzy attacks, but tailored to vehicular networks

Gear

Injects fake messages related to drive gear information, misleading the vehicle’s control system

Targets the CAN protocol, exploiting its lack of security features

Similar to other spoofing attacks, but specific to drive gear information

RPM

Injects fake messages related to RPM information, misleading the vehicle’s control system

Targets the CAN protocol, exploiting its lack of security features

Similar to other spoofing attacks, but specific to RPM information