Presenters

Hakeem WilsonFollow

Disciplines

Digital Communications and Networking

Abstract (300 words maximum)

Vehicle-to-Everything communications (V2X) is gaining additional ground as an upcoming ad hoc safety network. In V2X, basic safety messages are used for exchanging critical information between vehicles at a set broadcast rate. However, jamming attacks on the safety spectrum could deny V2X radios the ability to save lives on the roadway. This preliminary work analyzes two types of primitive jamming attacks performed on target V2X devices. Lab results reveal that V2X networks are easily susceptible to jamming attacks, due to all V2X standards lacking a requirement to detect/mitigate jamming. To avert this threat and promote safety of life on the roadways, we demonstrate a supervised machine learning model implemented at the baseband chipset could detect and classify the type of jamming attack with outstanding stability and an accuracy of 99.84%.

Academic department under which the project should be listed

SPCEET - Engineering Technology

Primary Investigator (PI) Name

Billy Kihei

Share

COinS
 

Detecting Primitive Jamming Attacks using Machine Learning in Vehicle-to-Everything Networks​

Vehicle-to-Everything communications (V2X) is gaining additional ground as an upcoming ad hoc safety network. In V2X, basic safety messages are used for exchanging critical information between vehicles at a set broadcast rate. However, jamming attacks on the safety spectrum could deny V2X radios the ability to save lives on the roadway. This preliminary work analyzes two types of primitive jamming attacks performed on target V2X devices. Lab results reveal that V2X networks are easily susceptible to jamming attacks, due to all V2X standards lacking a requirement to detect/mitigate jamming. To avert this threat and promote safety of life on the roadways, we demonstrate a supervised machine learning model implemented at the baseband chipset could detect and classify the type of jamming attack with outstanding stability and an accuracy of 99.84%.