Research Mentor: Zhijie Shi, Kaleel Mahmood, and Shengli Zhou
Secure wireless communications are the vital part of pervasive computing paradigm, especially important in the emerging Internet of Things (IoT) systems with billions of connected devices. The challenges are to provide reliable and secure wireless communications among computing devices in harsh environments and in the presence of adversaries. We plan to involve REU students in the following research tasks.
Secure and efficient physical layer in wireless communications: Security and energy efficiency are among the most important design goals in wireless communications. Malicious devices may pretend to be legitimate nodes and try to participate in system communications. Although upper layer protocols can have security mechanisms to verify the authenticity of messages, a device has to pass the received messages through many layers of protocol stacks from the physical layer. A more efficient way is to incorporate security mechanisms in the physical layer. Legitimate nodes can spend smallest amount of resources on messages from malicious nodes. One idea to encode a secure code in the chirp preamble, the signal that wakes up the communication module of devices to receive incoming data. The chirp signals can be formed with signals at many different frequencies. A sender can pick the frequencies based on a shared secret with the receiver. The receiver checks if a correct set of frequencies is present in the chirp, and only proceeds to receive and decode data signals if the chirp is correct. Not only more secure, the proposed method is also more energy efficient.
Countermeasures to jamming attacks: Wireless communications are more vulnerable to jamming attacks as adversaries can easily access the channels in IoT applications. Providing reliable communications under jamming attacks is critical to the function of many systems deployed in the field. Building upon our past research, we will further explore research topics including mechanisms and methods to prevent the attacks, detect the attacks, and respond to the attacks. To prevent attacks, we will experiment with spectrum-agile modulations. To detect the attacks, the computing devices will monitor the channel and use distributed algorithms to collectively detect the onset of attacks. When responding to the attacks, the devices can switch to modulations that are more robust.
Key establishment based on wireless channel properties: Key establishment is the foundation of many security protocols in sensor networks. We focus on key establishment based on physical layer mechanisms, where the randomness in the environments is leveraged to establish shared keys among devices. In particular, neighbouring nodes in a network can establish shared secrets from the unique channel properties that depend on their locations and propagation geometries.
REU students will have a chance to learn and understand the challenges in wireless communications. Not only exploring new ideas and solutions, they need to apply knowledge and skills to solve problems in real world applications;see an example REU project we have just supervised in Summer 2019 where students learn the communication modules, boot loaders, and security engines in practical sensor nodes from Texas Instruments. The proposed REU projects will be implemented and demonstrated on two different platforms: (a) software defined radio platforms, e.g., the universal software radio peripheral (USRP) units from National Instruments and the ADALM-PLUTO units from Analog Devices; and (b) in-air acoustic platforms with speakers and microphones. GNU Radio programming framework will be adopted, as we can leverage rich online resources while at the same time we can feed back our contributions to the research community quickly.