CuT-OFF SDR: a Cross-TechnOlogies Framework for Fingerprinting SDR

 

 

Founding Entity: Qatar National Research Fund (QNRF)

Amount of the Founding: $ 10,000.00

Duration of the Project: 1 year

Lead Investigator: Dr. Savio Sciancalepore

 

With the advent of wireless technology in early 2000s, a huge variety of devices started invading the market. They were equipped with cheap radio modules, specifically designed to enable transmission and reception operations at a given (narrow) set of communication frequencies, through dedicated and even more advanced modulation techniques.

While the legacy trend, especially in commercial systems, still consists in producing a dedicated hardware and specific devices for each usage and/or communication technology, starting from 2005 specific industries started developing general-purpose hardware, having the capability to implement in software some of the hardware-specific procedures and operations of the commercial devices. These new class of devices, namely Software Defined Radios (SDRs), are able to work on a pre-defined frequency band that is times more than the classical commercial devices. In addition, they are usually equipped with more powerful general-purpose processing units. In this way, it is possible to tune and program the software that runs on these SDRs to mimic the operations of a given radio device related to a certain wireless technology, thus making the device a legitimate network node.  Nowadays, a lot of SDRs are available on the market, with prices ranging from few tens up to few thousands of dollars. In addition, even more researchers, passionate and freelancers are dedicating time and efforts to improve freeware and open-source operating systems dedicated to SDRs, such as GNURadio.

From one side, SDRs have the potential of reducing costs of deployment of a new wireless technology, being devices that are natively ready to support whatever frequency targeted for the communication and the necessary computational and storage capabilities to carry out the necessary preparation and parsing of packets and related wireless signals. From the opposite side, SDRs represent a clear threat to any critical wireless infrastructure, being the classical dual use technology that can be used by malicious entities to easily break into wireless networks without being equipped with legitimate devices. An adversary that is able to implement via software the operations of a legitimate protocol could easy join the network and potentially disrupt its operations in a limited time, without the possibility to be detected as an external intruder. In parallel, in the last years, a research area named “Wireless Radio Fingerprinting” has started emerging. Contributions in this direction highlight the possibility to uniquely identify a given wireless device by looking at its imperfections and non-idealities in transmitted signals, i.e. the phase of the wireless signals, the fluctuation of the nominal frequency of the signals, modulation errors, magnitude errors, IQ offsets, and so forth.

Wireless radio fingerprinting is even more attracting the attention of researchers and industries, especially from the cybersecurity perspective, because of the possibility to authenticate at the physical layer the devices that join the network. This feature can provide fundamentally superior performance than traditional higher-layer solutions based on cryptography primitives.

To this aim, the current project intends to develop a framework that allows for the precise identification of any SDR at the physical layer of a wireless network, by resorting to wireless fingerprinting techniques. The framework will be developed by using GNURadio, that is the widely acceptable open-source and free-ware operating system that runs over a multitude of Software Defined Radios. As regarding the software development part of the project, the research will leverage the SDRs in possession of the Cybersecurity Lab of the HBKU-CSE, and will be developed in the spaces and offices already occupied by students and researchers of the Cybersecurity lab. As regarding the measurement part, the project will leverage measurement devices and tools available at the electronic measurement lab at HBKU-CSE, a recently established hands-on laboratory equipped with state-of-the art measurement devices such as oscilloscopes, spectrum analyzers, network analyzers, and so on.