Global Positioning System (GPS) technology is widely used in personal and industrial applications to acquire precise timing and positional information. However, its open-standard signals are vulnerable to spoofing attacks, which can cause serious damage if undetected. Employing detection methods is crucial in critical applications. Machine Learning (ML) methods have been successfully applied for spoofing detection, typically performing detection on individual samples. This work proposes a framework that takes a multivariate time-series window as input, enabling the neural network model to extract meaningful temporal information from the sample window for improved detection performance. We train a Convolutional Variational Autoencoder model using spoof-free samples under a representation learning framework. The detector's performance is evaluated using the publicly available TEXBAT dataset and simulated datasets. Our results show that the proposed detector achieves a True Positive Rate (TPR) above 99% for a low False Positive Rate (FPR) of 2% in both static and dynamic attack scenarios. Additionally, for the sophisticated attack scenario (DS-7) in the TEXBAT dataset, our detector achieved a TPR of 89% for an FPR of 3%, highlighting its robustness against different types of spoofing attacks.

Accurate load forecasting is crucial for energy management, infrastructure planning, and demand-supply balancing. The availability of smart meter data has led to the demand for sensor-based load forecasting. Conventional ML allows training a single global model using data from multiple smart meters requiring data transfer to a central server, raising concerns for network requirements, privacy, and security. To alleviate this issue, we propose a split learning-based framework for load forecasting. We split a deep neural network model into two parts, one for each Grid Station (GS) responsible for an entire neighbourhood's smart meters and the other for the Service Provider (SP). Instead of sharing their data, client smart meters use their respective GSs' model split for forward passes and only share their activations with the GS. Under this framework, each GS is responsible for training a personalized model split for their respective neighbourhoods, whereas the SP can train a single global or personalized model for each GS. Experiments show that the proposed models match or exceed a centrally trained model's performance and generalize well. Privacy is analyzed by assessing information leakage between data and shared activations of the GS model split.

Recent advances in IoT, machine learning, and edge computing have driven transformative paradigms like smart cities, grids, healthcare, and transportation systems, providing efficient solutions. This has led to a pervasive proliferation of connected devices, ranging from high-power computers to low-power sensors. Yet, the complex IoT architecture poses numerous vulnerabilities, demanding robust security measures. Existing firmware attestation techniques often encounter obstacles due to proprietary constraints, necessitating access to the device’s authentic firmware. To address this challenge, this paper proposes a novel software-based attestation framework that utilizes RAM traces from IoT devices for remote verification. By employing deep learning models trained in a representation learning paradigm, our framework empowers the remote verifier to authenticate the internal state of IoT devices. Leveraging data collected from real-world prototype devices, our approach achieves an impressive 100% detection rate for critical attacks on IoT devices with a false positive rate of 10−3. Remarkably, our framework preserves device availability and maintains low authentication latency, highlighting its efficacy and practicality for securing IoT ecosystems.

Privacy concerns have become increasingly prominent as machine learning (ML) models are adopted in an increasing number of sectors. The potential of unintended or malicious exposure of sensitive data, especially in E-Health solutions, has increased as these models are shared and deployed more broadly. In order to highlight the important problem of property inference attacks, which can result in privacy and data confidentiality breaches, this study focuses on inferring global characteristics of the underlying datasets used to train the ML models. Building upon the intriguing work by Ateniese et al. on property inference attacks on ML models, we present a novel property inference attack using Variational Auto-Encoders (VAEs). VAEs offer a strong answer to the difficult problem of inferring dataset attributes because of their reputation for being successful in modeling complex data distributions and producing synthetic data samples. Experiments on three healthcare and the US census datasets show that the proposed attack can effectively reveal underlying patterns in the training dataset with up to 94.29% accuracy. A comparison with the popular meta-classifier based property inference attacks shows that the proposed attack not only has better success rate, but can do so with half training data and a smaller number of shadow models.

With the proliferation of 4G and 5G mobile networks in smart cities, the adoption of Internet of Things (IoT) devices has surged, emphasizing the critical need for robust security measures. Existing firmware attestation techniques often require high computational budget or access to the device’s authentic firmware, posing challenges due to resource and proprietary constraints. To counter these two fundamental challenges, this article introduces a novel software-based attestation framework utilizing RAM traces from IoT devices for remote verification. In the proposed framework, the need for an authentic firmware copy is eliminated, and the most computationally intensive task is assigned to the gateway node of the IoT ecosystem. This approach yields a robust and highly accurate device attestation strategy, while imposing minimal computational demands on the verification device itself. Employing deep learning models trained in a representation learning paradigm, our framework enables the remote verifier to authenticate the internal state of IoT devices. Leveraging data collected from real-world prototype devices, under eight different applications, our approach achieves a remarkable 100% accuracy in detecting critical attacks on IoT devices with a false positive rate of$10^{-3}$. Notably, our framework preserves device availability and maintains low authentication latency, underscoring its efficacy and practicality for securing IoT ecosystems.

The reliable operation of phasor measurement units (PMU) in modern power grid monitoring system like wide-area measurement systems (WAMS) relies on accurate time synchronization, which is provided by the Global Positioning System (GPS). However, the open nature of civilian GPS signals makes PMUs vulnerable to time synchronization attacks (TSA), where attackers manipulate PMU time stamps by transmitting deceptive GPS signals near the PMUs. In this paper, we propose a framework for detecting TSA on PMUs using machine learning (ML) methods. We evaluate five ML algorithms, including Support Vector Machines, Random Forest, K-Nearest Neighbors, Gradient Boost, and Artificial Neural Network, and select seven complementary features that can be computed at the radio frequency (RF) and tracking stages of any commercial GPS receiver. Our detection protocol stands out from other similar ML-based methods in terms of speed, as it does not rely on waiting for the PVT solution. The Texas Spoofing Test Battery (TEXBAT) dataset is used to evaluate the proposed framework. We demonstrate that the ML models can effectively detect GPS spoofing with up to 99.9 % probability while maintaining less than 0.5 % false alarm and mis-detection probabilities. By providing early detection of GPS spoofing attacks on PMUs, the proposed framework has the potential to enhance the cybersecurity of WAMS.

In this paper, a robust sequential dictionary learning (DL) algorithm is presented. It is obtained by using a robust loss function in the data fidelity term of the DL objective instead of the usual quadratic loss. The proposed robust loss function is derived from the α-divergence as an alternative to the Kullback-Leibler divergence which leads to a quadratic loss. Compared to other robust approaches, the proposed loss has the advantage of belonging to class of redescending M-estimators, guaranteeing inference stability for large deviation from the Gaussian nominal noise model. The algorithm is derived via adaptive sequential penalized rank-l matrix approximation using a block coordinate descent approach to obtain the vector pairs of different rank-1 matrices. Performance comparison with similar robust DL algorithms on digit recognition highlights efficacy of the proposed algorithm.

Dictionary learning algorithms have been successfully applied to a number of signal and image processing problems. In some applications however, the observed signals may have a multi-subpsace structure that enables block-sparse signal representations. Based on the observation that the observed signals can be approximated as a sum of low rank matrices, a new algorithm for learning a block-structured dictionary for block-sparse signal representations is proposed. It's derived via sequential penalized low rank matrix approximation, where a block coordinate descent approach is used to estimate the matrix pairs that form the different low rank matrix approximations. Experimental results on synthetic and standard gray-scale images illustrating the performance of the proposed algorithm are provided.

In this paper, a robust sequential dictionary learning (DL) algorithm is presented. The proposed algorithm is motivated from the maximum likelihood perspective on dictionary learning and its link to the minimization of the Kullback-Leibler divergence. It is obtained by using a robust loss function in the data fidelity term of the DL objective instead of the usual quadratic loss. The proposed robust loss function is derived from the α-divergence as an alternative to the Kullback-Leibler divergence, which leads to a quadratic loss. Compared to other robust approaches, the proposed loss has the advantage of belonging to class of redescending M-estimators, guaranteeing inference stability from large deviations from the Gaussian nominal noise model. The algorithm is obtained by solving a sequence of penalized rank-1 matrix approximation problems, where the ℓ1-norm is introduced as a penalty promoting sparsity and then using a block coordinate descent approach to estimate the unknowns. Performance comparison with similar robust DL algorithms on digit recognition, background removal, and gray-scale image denoising is performed highlighting efficacy of the proposed algorithm.

In recent years, data driven methods have been successfully used for analyzing multi-subject functional magnetic resonance imaging (fMRI) datasets. These methods attempt to learn shared spatial activation maps (SM) or voxel time courses (TC) from temporally or spatially concatenated fMRI datasets respectively. Most of the methods proposed so far do not distinguish whether a particular SM/TC is a group level component or only present in a certain subject dataset. In this paper we present a new two stage algorithm which aims to separate the joint and sub-specific information from the temporally concatenated multi-subject datasets. The proposed method is based on the singular value decomposition (SVD) and penalized rank-one matrix approximation. Simulation experiments are used to demonstrate this ability of the proposed algorithm followed by validation on real experimental task-fMRI datasets.

In recent history, dictionary learning (DL) methods have been successfully used for analyzing multi-subject functional magnetic resonance imaging. These algorithms try to learn group-level spatial activation maps (SM) or voxel time courses (TC) from temporally or spatially concatenated fMRI datasets respectively. However, in multi-subject fMRI studies, we are interested in both group-level TCs as well as SMs. In this paper, we propose a DL algorithm which combines temporally and spatially concatenated fMRI datasets to learn not only the shared TC/SM pairs but also the subject-specific ones. We do this by separating group-level information and sub-specific information from each subject fMRI dataset. Performance of the proposed algorithm is illustrated using simulated as well as experimental task fMRI datasets.

Multiple-set canonical correlation analysis (mCCA) is a generalization of canonical correlation analysis (CCA) to three or more sets of variables. It aims to study the relationships between several sets of variables and it subsumes a number of interesting multivariate data analysis techniques as special cases. The quality and interpretability of the mCCA components are likely to be affected by the usefulness and relevance of each set of variables. Therefore, it is an important issue to identify each set of significant variables that are active in the relationships between sets. In this paper mCCA is extended to address the issue of variable set selection. Specifically a block sparse multiple set canonical correlation analysis (BSmCCA) algorithm is proposed to combine mCCA with ℓ2-norm type penalty in a unified framework. Within this framework sets of variables that are not necessarily relevant are removed. This makes BSmCCA a flexible method for analyzing for Multi-Subject functional magnetic resonance imaging (fMRI) data sets. The performances of the proposed BSmCCA algorithm are illustrated through on block design paradigm finger taping fMRI datasets.

Sequential dictionary learning algorithms has gained widespread acceptance in functional magnetic resonance imaging (fMRI) data analysis. However, many problems in fMRI data analysis involve the analysis of multiple-subject fMRI data sets and the existing algorithms do not extend naturally to this case. In this paper we propose an algorithm dedicated to multiple-subject fMRI data analysis. The algorithm is named SMSDL for sequential multi-subject dictionary learning and differs from existing dictionary learning algorithms in its dictionary update stage. This algorithm is derived by using a variation of the power algorithm in the dictionary update stage to extract the common information among the multiple-subject fMRI data sets. The results of the proposed dictionary learning algorithm is a set of time courses which are common to the whole group of subjects and an individual spatial response pattern for each of the subjects in the group. The performance of the proposed algorithm are illustrated through a simulation and an application on real fMRI datasets.

Sequential dictionary learning via the K-SVD algorithm has been revealed as a successful alternative to conventional data driven methods, such as independent component analysis for functional magnetic resonance imaging (fMRI) data analysis. fMRI data sets are however structured data matrices with notions of spatio-temporal correlation and temporal smoothness. This prior information has not been included in the K-SVD algorithm when applied to fMRI data analysis. In this paper, we propose three variants of the K-SVD algorithm dedicated to fMRI data analysis by accounting for this prior information. The proposed algorithms differ from the K-SVD in their sparse coding and dictionary update stages. The first two algorithms account for the known correlation structure in the fMRI data by using the squared Q, R-norm instead of the Frobenius norm for matrix approximation. The third and last algorithms account for both the known correlation structure in the fMRI data and the temporal smoothness. The temporal smoothness is incorporated in the dictionary update stage via regularization of the dictionary atoms obtained with penalization. The performance of the proposed dictionary learning algorithms is illustrated through simulations and applications on real fMRI data.

Sequential dictionary learning algorithms have been successfully applied to functional magnetic resonance imaging (fMRI) data analysis. fMRI data sets are, however, structured data matrices with the notions of temporal smoothness in the column direction. This prior information, which can be converted into a constraint of smoothness on the learned dictionary atoms, has seldomly been included in classical dictionary learning algorithms when applied to fMRI data analysis. In this paper, we tackle this problem by proposing two new sequential dictionary learning algorithms dedicated to fMRI data analysis by accounting for this prior information. These algorithms differ from the existing ones in their dictionary update stage. The steps of this stage are derived as a variant of the power method for computing the SVD. The proposed algorithms generate regularized dictionary atoms via the solution of a left regularized rank-one matrix approximation problem where temporal smoothness is enforced via regularization through basis expansion and sparse basis expansion in the dictionary update stage. Applications on synthetic data experiments and real fMRI data sets illustrating the performance of the proposed algorithms are provided.