Trojans represent a severe threat to hardware security and trust. This work investigates the Trojan detection problem from unique viewpoint proposes novel localization method targeting FPGA netlists. The proposed automatically extracts rich structural behavioral features at look-up-table (LUT) level train an explainable graph neural network (GNN) model for classifying design nodes in netlists identifying Trojan-infected ones. Experimental results using 183 benchmarks show that our...

This work proposes a novel hardware Trojan detection method that leverages static structural features and behavioral characteristics in field programmable gate array (FPGA) netlists. Mapping of design sources to look-up-table (LUT) networks makes these explicit, allowing automated feature extraction further effective through machine learning. Four-dimensional are extracted for each signal random forest classifier is trained net classification. Experiments using Trust-Hub benchmarks show...

Taint-propagation and X-propagation analyses are important tools for enforcing circuit design properties such as security reliability. Fundamental to these effective models accurately measuring the propagation of information calculating metadata. In this work, we formalize a unified model reasoning about taint- behaviors verifying related behaviors. Our developed from perspective flow can be described using standard hardware description language (HDL), which allows formal verification both...

Masking is a commonly used countermeasure for protecting cryptographic implementations from power side channel analysis. Rotating S-Box masking (RSM) state-of-the-art technique, whose effectiveness has been testified in ongoing worldwide DPA contest. However, does not provide sufficient protection against active attacks through fault injection. In this paper, we propose simple yet effective correlation attack on two generations of RSM AES schemes. We demonstrate mathematical formulations...

Neural networks have been successfully applied in numerous domains with the help of high-quality training samples. However, datasets containing noises and outliers (i.e., corrupted samples) are ubiquitous real world. When using these as samples, most neural exhibit poor predictive performance. In this paper, motivated by modal regression, we propose a Modal Network, which is robust to Specifically, network can reveal likely trends samples without overfitting On theoretical side, establish...

Recently, adversarial metric learning has been proposed to enhance the robustness of learned distance against perturbations. Despite rapid progress in validating its effectiveness empirically, theoretical guarantees on and generalization are far less understood. To fill this gap, paper focuses unveiling properties by developing uniform convergence analysis techniques. Based capacity estimation covering numbers, we establish first high-probability bounds with order O(n^{-1/2}) for pairwise...

Abstract Large-scale Agent data partitioning is the premise of parallel distributed computing in process ABMS (Agent-based Modeling and Simulation) . Based on distance-based K_medoids clustering algorithm, this paper proposes an improved algorithm (DensityRepel-K_medoids), which implemented by high-performance programming language X10 applied to large-scale simulation based distance interaction. The DensityRepel-Kmedoids first determines density value repulsion each set, secondly pre-selects...

Consortium blockchain has been widely used in different scenarios, where members demand that their uploaded data could be audited under identities without exposing the themselves. However, so far, no solution of privacy-preserving auditing proposed. To address problem, we propose zkrpChain, which focuses on protection integrity and privacy by while leaving public. In is based Hyperledger Fabric Bulletproofs, both standard-range arbitrary-range zero-knowledge range proofs generation...

Symmetric cryptographic functions are known to leak secret information through power side channels. Although masking provides an effective mitigation such leakage, these protection designs vulnerable advanced side-channel analysis based on machine learning. In this work, we propose a framework leverage ensemble learning for masked AES implementation. Before the key recovery, use techniques including decision tree, random forest, BP nerual network, and recover masks. We also implement data...

Graphics Processing Unit (GPU) performs graphics computing and its architecture has developed from the fixed function pipeline to programmable unified pipeline. Unified promises dynamic load balancing guarantees high parallel of GPU. This paper presents design implementation a The shader is based on SIMD SIMT architecture. On thread level, full-load capability by managing scheduling. instruction controls execution hardware unit. We finish algorithm, Verilog RTL implementation. verification...

Satisfiability don't cares (SDCs) have recently been exploited for malicious design modifications, where an SDC signal pair is used as two discrete Trojan triggers. The care condition can never be satisfied under normal operation and the only activated through fault injection. However, each trigger able to switch normally. Consequently, Trojans are resilient state-of-the-art side channel well switching probability analysis based detection techniques. In this paper, we propose a novel method...

Software attacks that exploit the hardware security vulnerabilities of processors have become new breakthrough points for hackers, which pose severe threats to and trust. This paper proposes a novel system verification method based on instruction set architecture (ISA) level information flow tracking (IFT), is capable modeling checking properties in both software designs. We use RISC-V as demonstration, lately developed open-source ISA widely used Internet Things facing due universal...

Third-party intellectual property cores are essential building blocks of modern system-on-chip and integrated circuit designs. However, these design components usually come from vendors different trust levels may contain undocumented functionality. Distinguishing such stealthy lightweight malicious modification can be a challenging task due to the lack golden reference. In this work, we make step towards for assurance by developing method identifying preventing hardware Trojans, employing...

Trojan horses represent a major threat to hardware security and trust. In this work, we propose novel detection method based on explainable graph neural networks (GNNs) targeting FPGA netlists. We leverage the rich explicit structural features behavioral characteristics at LUT, which offers an ideal abstraction level granularity for detection. A GNN model with optimized class-balanced focal loss is trained automated feature extraction classification. Based Granger causality theory, develop...

In this letter, we consider the proof of non-stationary channel polarization theory. First construct a multi-channel stochastic process for operation. Then based on process, extend Ar{\i}kan's standard martingale method average capacity and Bhattacharyya parameter, by which have proved

High performance 3D graphics processor is one of the most important hardware modules in modern mobile SoC design. This paper presents design a fully programmable taking advantage unified shader architecture for applications. We adopt multi-thread scheduling to utilize data, instruction parallelism and reduce data dependency stalls. In core design, we single multiple (SIMD) support four concurrent scalar operation. Moreover, propose novel special function unit floating point transcendental...