Real-Time Detection of Power Analysis Attacks by Machine Learning of Power Supply Variations On-Chip

Power side channel attack analysis and detection

“…This section compares the proposed SCA sensor with ML-based techniqus [7] and [8] (Table 2) Approach: The ML based detection techniques create an SCA aware environment during test phase of the chip to train the LR classifier. We exploit the inherent RO frequency shift under SCA which serves as a real-time detection metric; (ii) Overhead: The silicon area of the proposed work is 0.044% and 0.1065% than that of [7,8], respectively.…”

“…GE is the area ofNAND2 gate equivalent. ADCs contribute to more than 90% of the area and power of the ML based [7,8] may suffer from these factors;(iv) Test overhead: ML-based techniques [7,8] require large data sets of power signatures to be sampled and stored in the memory. These data sets are then used to train the LR models.…”

“…Storage overhead: The chip's power trace and Nc1ean values need to be stored in the memory in ML-based techniques [7,8] while this is not required for the proposed reference-free approach; (vi) Scalability of detection: The proposed method is scalable to resistances lower than 1 Q provided the detection window is increased (incurs area and power overheads).…”

“…The existing methodologies for SCA detection rely on Machine Learning (ML) [7,8]. For example, the change in the power grid impedance due to the insertion of a SCAR at the power supply is leveraged for detection [7]. The chip's power profiles with and without SCARS of various values are captured during the design phase.…”

Implementation of Decision Tree Algorithm to Classify Knowledge Quality in a Knowledge Intensive System

“…Unsupervised learning focuses at identifying patterns in a dataset without known experience or samples [19,20]. Common supervised learning algorithm are Artificial Neural Network, Decision Tree, Linear Regression, Logistic Regression, K-Nearest Neighbour, Naïve Bayes , Random Forest and Support Vector Machine [49][50][51]. Common unsupervised learning algorithm are Apriori, Equivalence Class Transformation, Expectation Maximisation, Frequent Pattern-Growth, Hierarchical Clustering, K-Means Clustering, Mean Shift and Spectral Clustering [49,52].…”