For more updated news on our work on signal prcessing, power systems, sensors, and bandwidth extension, please follow our Google Scholar link.
Encryption of Analog Signal
One of the main reasons for CPS vulnerability is the legitimate analog signal, which is going to be measured by the sensor, is not encrypted before going into the transducer. Therefore, the attacker can use a fake signal to corrupt the legitimate signal. This problem can be solved by encrypting the legitimate analog signal with a key in the analog domain and decrypting the legitimate signal on the transducer side using the same key. My research explores analog domain encryption. This is an open domain and our research will be pushing along this direction. We are using different noise padding and cryptographic techniques. According to preliminary results, our methods have promising data obfuscation capabilities.
Distributed Sensor Architecture for CPSs and IoTs
Today, data is generated with higher velocity and higher volume than can be feasibly stored, raising the need for new algorithms, software abstractions, and systems. Majority of these data are coming from different sensors in CPSs and IoTs. However, today’s CPS architecture is missing distributed sensor systems which will help to provide distributed online data processing. The idea of distributed online data processing will be accomplished inside of the distributed sensor systems enabling less data handling by the main system controller. This will enable comparatively low-power and low-complexity algorithms running inside distributed sensor systems. A viable option can be using machine learning (ML) algorithms to create an appropriate context and abstraction of the sensor data from distributed sensor systems. Therefore, during an attack, the sensor data can be recovered from a proper context using abstracted sensor data. The data abstraction can be made intelligent and adaptive to tackle the continuous change of the sensor environment.
Critical Infrastructure Design
Our research sits at the intersection of hardware and software layers in Critical Infrastructures, exploring how interactions of cyber and physical components open a “Pandora’s Box” of unknown threats, specifically about how they affect the safety and controllability of closed-loop control from sensing to actuation. Critical Infrastructures are engineered systems that are built from, and depend upon the seamless integration of computation and physical components. Most attacks on Critical Infrastructure can be propagated from the physical domain to the cyber domain or vice-versa and hence, can be termed as cross-domain attacks. To understand these cross-domain attacks, a very different set of methodologies and tools are needed. Moreover, as these cross-domain attacks involve hardware and software layers, defenses against these vulnerabilities also demand new hardware/software co-design approaches to detect, contain and isolate vulnerabilities in CPSs and IoTs. My recent and ongoing studies investigate technological threats to the safety of Critical Infrastructures. Themes I also investigate include attack forensics and provenance, hardware and microarchitectural security, machine learning in Critical Infrastructures, low-power hardware and algorithms for robust defense, and distributed data architecture in Critical Infrastructures.