7 January 2019 in IEEE Transactions on Aerospace and Electronic Systems
This paper proposes a deep learning-based intelligent method for detecting and identifying radio signals considering two applications: (1) cognitive radar for identifying micro unmanned aerial systems (micro-UASs) and (2) an automated modulation classification (AMC) scheme for cognitive radio (CR), which can be used for aeronautical communication systems. Our proposed intelligent method is designed of a spectral correlation function (SCF) based feature extractor and a low-complexity deep belief network (DBN) classifier with low FPGA logic utilization.
30 March 2020 in IEEE Transactions on Emerging Topics in Computing
Thanks to the advances in machine learning, data-driven analysis tools have become valuable solutions for various applications. However, there still remain essential challenges to develop effective data-driven methods because of the need to acquire a large amount of data and to have sufficient computing power to handle the data. In many instances these challenges are addressed by relying on a cloud computing vendor. However, although commercial cloud vendors provide valuable platforms for data analytics, they can suffer from a lack of transparency, security, and privacy-preservation. Furthermore, reliance on cloud servers prevents applying big data analytics in environments where the computing power is distributed. To address these challenges, a decentralized, secure, and privacy-preserving computing paradigm is proposed to enable an asynchronized cooperative computing process amongst distributed and untrustworthy computing nodes that may have limited computing power and computing intelligence. This paradigm is designed by exploring blockchain, decentralized learning, and homomorphic encryption techniques. The performance of the proposed paradigm is evaluated via different scenarios in the simulation section.
18 June 2018 in IEEE Journal of Selected Topics in Signal Processing
This paper proposes an attack-resilient intelligent soft computing based security control for energy management architecture for a hybrid emergency power system of More-Electric Aircrafts (MEAs). Our proposed architecture develops an Adaptive Neuro-Fuzzy Inference System (ANFIS) -based method in conjunction with a specific Recurrent Neural Network architecture called Long Short-Term Memory (LSTM) to evaluate the integrity of the power output of the critical components which have higher vulnerability to the potential cyber-attacks and critical for the effective energy management and emergency control. We evaluate the integrity of the critical measurements that have high vulnerability to the potential cyber-attacks by using LSTM techniques. After identifying the compromised measurements, we continue to activate the closed-loop ANFIS mechanism in which the structure of ANFIS is controlled according to the accuracy of its current estimations by probing the physical couplings in the system. In our simulation, we evaluate the performance of our proposed LSTM-ANFIS -based method to support the integrity of the Energy Management Strategies (EMSs) used in hybrid emergency power system for more-electric aircrafts by using TensorFlow and MATLAB/Simulink co-simulation environment. Our simulation results illustrate the effectiveness of our proposed method in effectively evaluating the integrity of critical data and achieving resilient control.
21 December 2023 in IEEE Transactions on Education
Contribution: A novel proactive and collaborative learning paradigm was proposed to engage learners with different backgrounds and enable effective retention and transfer of the multidisciplinary artificial intelligence (AI)-cybersecurity knowledge. Specifically, the proposed learning paradigm contains: 1) an immersive learning environment to motivate the students for exploring AI/machine learning (ML) development in the context of real-world cybersecurity scenarios by constructing learning models with tangible objects and 2) a proactive education paradigm designed with the use of collaborative learning activities based on game-based learning and social constructivism. Background: Increasing evidence shows that AI techniques can be manipulated, evaded, and misled, which can result in new and profound security implications. There is an education and training gap to foster a qualified cyber-workforce that understands the usefulness, limitations, and best practices of AI technologies in the cybersecurity domain. Efforts have been made to incorporate a comprehensive curriculum to meet the demand. There still remain essential challenges for effectively educating students on the interaction of AI and cybersecurity. Intended Outcomes: A novel proactive and collaborative learning paradigm is proposed to educate and train a qualified cyber-workforce in this new era where security breaches, privacy violations, and AI have become commonplace. Application Design: The development of this learning paradigm is grounded in the pedagogical approaches of technology-mediated learning and social constructivism. Findings: Although the research work is still ongoing, the prototype learning paradigm has shown encouraging results in promoting the learners’ engagement in applied AI learning.
6 May 2020 in IEEE Transactions on Circuits and Systems I: Regular Papers
The biological inner ear, or cochlea, is a sophisticated signal processing system that performs spectrum analysis over an ultra-broadband frequency range of ~20 Hz to 20 kHz with exquisite sensitivity and high energy efficiency. Electronic cochlear models, which mimic the exponentially-tapered structure of the biological inner ear using bidirectional transmission lines or filter cascades, act as fast and hardware-efficient spectrum analyzers at both audio and radio frequencies. This paper describes a cochlea-based digitally-programmable single-chip radio frequency (RF) spectrum analyzer in 65 nm CMOS. This “RF cochlea” chip includes a transmission-line active cochlear model with 50 parallel exponentially-spaced stages that analyzes the radio spectrum from 1.0-8.3 GHz. The outputs of all stages are encoded in parallel as delta-sigma (Δ-Σ) modulated digital signals for real-time demodulation and analysis by a digital back-end processor. The chip consumes 418 mW and typically generates ~1 GS/s of total data at an ENOB of 5-6 bits. An artificial intelligence (AI)-driven single-channel cognitive radio (CR) receiver based on the RF cochlea has also been implemented and tested.