Securing wireless communication systems at the physical layer

Marko Jacovic

doi:10.17918/00010640

With the emergence of new wireless technologies in every-day lives lives such as in Cyber-Physical Systems (CPS), Vehicular, Unmanned Aerial Vehicle (UAV) communications, and the Internet of Things (IoT), protecting devices against external threats is essential for society to thrive in the future. Wireless devices face both passive attacks, in which eavesdropping results in the theft of sensitive information, and active attacks which interfere with communication. Wi-Fi Protected Access (WPA2/3) has been shown to be vulnerable in the past, particularly for interconnected short-range IoT devices. Physical (PHY) layer security leverages existing components of the PHY layer to provide complementary techniques to strengthen existing security protocols at higher layers. Radio frequency (RF) jammers disrupt the wireless RF medium by transmitting signals with the malicious intent to interfere with legitimate communication systems. Defense against jamming attacks includes detection of a threat and implementation of mitigation techniques to attempt to circumvent the interference source. An inherent challenge exists when evaluating the impact of jamming as well as classification and mitigation techniques under complex emerging wireless communication scenarios. In these cases, the scenarios are difficult to examine with hardware in the loop due to mobility, zoning restrictions, scalability of nodes, repeatability of testing, and of course the legality of jamming the RF medium. Previous assessment of jamming defense techniques have focused predominately on simulation and analytical solutions, which simplify practical constraints associated with hardware and RF impairments, or small-scale indoor hardware experiments. Existing hardware implementation of mitigation techniques have consisted of low data-rate hostile military applications or have not provided general solutions for different jammer types. In this dissertation defense techniques for wireless communication systems against both passive and active attacks are presented. A method is introduced to secure wireless devices from eavesdropping attacks through the use of frequency induction at the PHY layer. Shared keys are used by a pair of legitimate devices to map frequency shifts applied to the payload of packets. Encryption and decryption techniques are analyzed in simulation and unit-level hardware implementation. A platform suitable for the evaluation of a diverse array of communication systems under the attack of various types and number of RF jammers is presented. The focal point of the jamming experimentation framework consists of a bridge connecting ray-tracing software to a highly scalable channel emulator which can be connected to either flexible software defined radio (SDR) or commercial off the shelf (COTS) wireless devices. Ray-tracing software is used to develop realistic site-specific channels which vary over time and are used to program the emulator, such that external hardware consisting of legitimate devices and jamming sources may be considered for analysis under complex scenarios. Case studies are provided to demonstrate the use of the framework for different applications and jamming threats. Furthermore, a real-time full-stack SDR-based platform that leverages cross-layer metrics, pattern reconfigurable antennas, and machine learning to mitigate active RF jamming attacks is presented. Pattern reconfigurable antennas provide directional power gain to reduce the effects of the active attack. Classification of jamming events is used to activate an antenna state selection algorithm. The mitigation technique is evaluated through both over-the-air and ray-tracing emulation experiments under various jamming scenarios.

Securing wireless communication systems at the physical layer

Files and links (1)

Abstract

Metrics

Details

Securing wireless communication systems at the physical layer

Files and links (1)

Abstract

Metrics

Details

Drexel University Social media