The CAN was designed to operate as a closed network system; that is$,$ all the communicating nodes are trusted. The CAN does not support security mechanisms to identify and isolate malicious nodes, such as authentication. This design was not a problem when automobiles were standalone entities; however, today, ITS applications communicate with external entities, such as other vehicles, in the case of CACC, which makes it possible for attackers to exploit vulnerabilities in these applications and attack the connected vehicles.

Cyber$-$attacks goal on connected vehicles is to inject CAN messages that aim to change the vehicle's behavior, such as increasing$/$decreasing the speed or disabling the break. You are given in the following datasets collected from the CAN bus of a vehicle:

In normal driving.

Under spoofing the speed reading $($by injecting fabricated speed CAN messages$).$

Under spoofing the RPM reading $($by injecting fabricated RPM CAN messages$).$