Efficient and automated wafer yield prediction is central to cost control and process optimization in intelligent semiconductor manufacturing. As the initial testing stage in wafer inspection, Wafer Acceptance Testing (WAT) data contain critical process-related information. However, its high dimensionality, redundancy, and nonlinear inter-dependencies pose significant challenges to conventional yield prediction models, such as high computational overhead and limited generalization capability. Moreover, existing approaches often lack an automated framework capable of jointly addressing feature redundancy and model complexity. This paper proposes a collaborative and automated prediction framework that integrates feature selection and machine learning. First, an enhanced Binary Zebra Population Optimization Algorithm (BZPOA) is introduced, which incorporates redesigned exploration and development mechanisms to automatically identify key feature subsets from high-dimensional parameters, substantially reducing data redundancy and computational dimensionality. Second, a Bayesian hyperparameter-optimized XGBoost model is constructed, utilizing the Tree-structured Parzen Estimator (TPE) to achieve deep co-optimization of model parameters and feature space, thereby overcoming the inefficiency and overfitting issues commonly associated with manual parameter tuning. Experiments on a real-world dataset demonstrate that the proposed framework achieves average Recall, Precision, and F1-scores of 0.872, 0.917, and 0.902, respectively. Compared with the full-feature baseline, the BZPOA-selected feature subset improves predictive performance by 5%–10%, attains an AUC of 0.904, and significantly reduces per-wafer prediction time. Cross-factory transfer experiments further confirm the robustness of the proposed system.

Arbiter physical unclonable function (APUF) is a hardware security primitive that generates security keys by utilizing unavoidable process variations during chip manufacturing. However, the structure based on linear additive function makes it vulnerable to machine learning (ML) attacks. This paper proposes a parallel feedback obfuscation PUF (PFO PUF) design, which uses intermediate arbitration signals of the lower-layer APUF to generate the hidden challenge of the upper-layer APUF, enhancing the overall nonlinearity of the structure. The obfuscation module makes weight judgment for intermediate arbitration signals of upper-layer and lower-layer APUFs, which obfuscates the real response of PUF. We further design a variant of PFO PUF called reconfigured challenge obfuscation PFO PUF (RPFO PUF) and propose its lightweight device authentication protocol. RPFO PUF enhances the resistance of the original PFO PUF against reverse engineering (RE) and improves its Strict Avalanche Criterion (SAC) characteristic by reordering the challenges and incorporating weak PUF responses. The proposed PFO PUF and RPFO PUF were comprehensively evaluated via Python-based simulations and FPGA measurements. In Python simulations, both designs show strong resistance to state-of-the-art ML attacks, with logistic regression (LR), support vector machine (SVM), and covariance matrix adaptation evolution strategies (CMA-ES) yielding near 50% prediction accuracies under various PUF configurations. Although deep neural network (DNN) achieves up to 69.52% prediction accuracy on the PFO PUF, it drops to ∼50% on the RPFO PUF. FPGA results further confirm this, with the (32, 11)-RPFO PUF achieving a maximum prediction accuracy of only 51.47% across all four ML attacks. Moreover, both designs incur low hardware overheads, requiring just 743 and 2145 gate equivalents (GEs), respectively.

Physical unclonable functions (PUFs) are critical hardware security primitives that extract entropy from intrinsic manufacturing process variations (MPVs) to generate unique cryptographic responses. They have demonstrated extensive application prospects in lightweight encryption and authentication for Internet of Things (IoT) devices. However, the classic arbiter PUF (APUF) is inherently vulnerable to machine learning (ML) modeling attacks due to the linear mathematical model of its delay circuits. To address this challenge, this article proposes a novel postprocessing obfuscation architecture named recursive Markov chain PUF (RMC PUF). The proposed architecture couples an APUF core with Markov obfuscation modules. By exploiting the intermediate stage responses of the APUF as an entropy source, the system constructs a 1-D recursive stochastic state transition mechanism. This approach recursively transforms and obfuscates the raw entropy based on Markov Chain theory, achieving a PUF system with stability and robust resistance against ML attacks. Through this efficient recursive obfuscation strategy, the design achieves precise utilization of the APUF’s intrinsic entropy. The work is validated through both numerical simulation and FPGA prototyping. Experimental results show that the prototype maintains a high reliability of 98.9%–99.9% within a voltage range of 0.8–1.2 V, and a reliability of 96.1%–99.7% within a temperature range from$- 10~{^{\circ }}$C to$80~{^{\circ }}$C. Comparative analysis indicates that the proposed RMC PUF significantly outperforms existing structures in security, limiting the prediction accuracy of four ML attack models [logistic regression (LR), artificial neural networks (ANNs), deep neural networks (DNNs), and covariance matrix adaptive evolutionary strategy (CMA-ES)] to no more than 54.72%.

Physical Unclonable Functions (PUFs) are crucial for lightweight authentication in the Internet of Things (IoT), but existing PUFs often have poor statistical properties and are vulnerable to machine learning attacks. Designs implementing the Strict Avalanche Criterion (SAC) lack sufficient theoretical foundation. This paper introduces quantitative metrics to evaluate the SAC performance of strong PUFs and conducts rigorous analysis on Arbiter PUFs (APUFs) and their classic variants, addressing imprecision in existing methods. Based on these metrics, we developed an algorithm to optimize the SAC performance of strong PUFs by adjusting the challenge sequences. This optimization improved the SAC performance of the 2-XOR APUF by 59% without additional resource consumption, making it comparable to the 4-XOR APUF; We also provided mathematical proof for the optimal solution. Furthermore, we propose the SAC Optimized Shuffled XOR Arbiter PUF (SOS XOR APUF), which improves SAC performance by 84% compared to the 3-XOR APUF with the same entropy source. It addresses the inherent defect of poor statistical properties when two adjacent bits in the challenge flip, achieving a theoretical response flip probability of 0.5. The SOS XOR APUF resists existing machine learning attacks---including Logistic Regression (LR), Covariance Matrix Adaptation Evolution Strategy (CMA-ES), Artificial Neural Network (ANN), and Deep Neural Network (DNN)---with prediction accuracy below 55%. Finally, we designed and verified an authentication protocol based on this PUF in the ProVerif environment, achieving mutual authentication between IoT devices and servers, preventing secret information from being stolen, and enhancing the security of the PUF structure.

Strong physical unclonable functions (PUFs) are promising circuits for lightweight Internet of Things (IoT) authentication and security. However, existing strong PUFs exhibit very low cryptographic nonlinearity (NL), making them vulnerable to machine learning (ML) modeling and cryptanalytic attack. To address this issue, we propose the Bent function PUF (BF PUF) based on Maiorana-McFarland (M-M) constructed Bent functions, which obfuscates the responses of the strong PUF to enhance resistance against modeling attacks. The core idea is to employ the M-M construction method for Bent functions to ensure maximum cryptographic NL to resist modeling attacks. A Feistel network is configured using weak PUF responses as keys to achieve device-specific and unpredictable mappings of input challenges while meeting the requirements of the M-M Bent function construction. A Python-based model of the BF PUF was developed, and simulation results indicate that the cryptographic NL of the proposed BF PUF outperformsk-xorarbiter PUFs (APUFs) (${k} =2$, 4, 6). The proposed BF PUF was also implemented and evaluated on the FPGA hardware platform. The experimental results show that under modeling attacks using four ML algorithms—logistic regression (LR), artificial neural networks (ANNs), deep neural networks (DNNs), and covariance matrix adaptation evolution strategies (CMA-ES)—the best prediction accuracy under these four modeling attack algorithms is 52.60%. The reliability under temperature fluctuations ranging from$- 10~^{\circ }$C to$80~^{\circ }$C is between 84.20% and 99.78%.

As a critical hardware security primitive for the authentication within the Internet of Things (IoT), the physical unclonable function (PUF) represents an innovative security design paradigm for integrated circuits. However, the linear challenge-response mapping of the arbiter PUF (APUF) and variants render these structures more susceptible to modeling attacks due to their delayed linear structure. In this article, we propose a nonlinearized demultiplexer time-to-digital converter (DEMUX-TDC) PUF. This PUF uses a quantized delay difference technique to alter the traditional response generation mechanism, demonstrating robust resistance against modeling attacks. First, the proposed PUF employs a segmented APUF variant structure configured in both front and back segmented modes to generate a source of delay difference entropy. Additionally, the scheme incorporates a multilayer differential tapped TDC circuit to quantize the delay differences into digital codes, followed by a linear feedback shift register (LFSR) to obfuscate the final output response. We further propose a highly secure mutual authentication protocol based on reconfigurable PUF, by leveraging the characteristics of front and back segments of the PUF’s challenge. Evaluation of the proposed scheme on the implementation of Xilinx Virtex-7 and Spartan-6 field-programmable gate array (FPGA) demonstrates that the uniqueness and uniformity can reach ideal value, in the condition of prediction accuracy across six modeling attacks remaining around 50%.

Physical Unclonable Functions (PUF) are important hardware security primitives used for generating keys and identity authentication, with wide applications in the Internet of Things security. However, the security of strong PUF faces serious threats from modeling attacks, especially in the case of Arbiter PUF and their variants that involve additive linear integration of entropy sources. This paper proposes a Ring-Oscillator-Integrated-Linear-Feedback-Shift-Register-based PUF (ROinLFSR PUF) that achieves immunity to modeling attacks by highly nonlinearly integrating independent responses from weak RO PUFs using a configurable LFSR. To increase the efficiency of entropy extraction in hardware resources, dual entropy sources extraction is performed on the period and duty cycle of RO. Python simulation and FPGA experimental results demonstrate that the proposed PUF has intrinsic resilience against modeling attacks. And the proposed PUF achieves good results in reliability, uniqueness, uniformity, and randomness.

Arbiter Physically Unclonable Functions (APUFs) are hardware security primitives that leverage manufacturing process variation to generate security keys. They can produce exponential challenge-response pairs (CRPs) with minimal hardware overhead. However, the symmetric nature of linear additive functions makes them vulnerable to modeling attacks rooted in machine learning. To address this issue, this paper introduces a novel design called Parallel Feed Obfuscation PUF (PFO PUF). In this approach, the intermediate decision signals from the lower APUF are used as a concealed challenge for the upper APUF, enhancing the overall nonlinearity of the dual-APUF. Additionally, obfuscation modules are employed to determine the weights of the intermediate decision signals from both the upper and lower APUFs, protecting the actual response. Experimental results demonstrate that the proposed PFO PUF effectively withstands four advanced machine learning attack algorithms, including Logistic Regression (LR), Support Vector Machine (SVM), Deep Feedforward Neural Network (DFNN), and Efficient CANDECOMP/PARAFAC Tensor Regression Network (ECPTRN). The prediction accuracy of these four algorithms is consistently below 66.30%. Compared with other enhanced structures based on APUF, PFO-PUF only uses 493 LUTs and has lower resource overhead.

The physical unclonable function (PUF) is a hardware security primitive that is used to generate secret keys or identity authentication for chips using random manufacturing process variation (MPV). The PUF based on ring oscillator (RO PUF) has been extensively studied in recent years because of its high robustness and ease of design. Although the performance has been optimized in previous studies, several uniqueness, reliability, and theoretical foundation concerns still remain. This article presents a transistor-level parameters-based quantitative theoretical model, which clearly reveals several design guidelines for improving the reliability of RO PUF. Furthermore, a PUF based on a feedback ring oscillator (RO) structure is proposed, which combined the RO topology and the drafting effect of XOR gates to enhance the uniqueness and reliability. The correctness of the theoretical model was verified by the SPICE simulation experiment result. And in the FPGA experiment result, the uniqueness and the reliability of feedback RO PUF using the same hardware resources on the same chip was superior to that of RO PUF. The theoretical research method of RO PUF used can be widely applied to other PUFs using ring topology and feedback RO PUF is a great substitute for RO PUF.

The physical unclonable function (PUF) is a hardware security primitive that can be used to prevent malicious attacks aimed at obtaining device information at the hardware level. The ring oscillator (RO) PUF has attracted considerable research attention. To improve the reliability of the RO PUF under voltage and temperature changes, the response of the duty-cycle (DC) PUF was obtained by comparing the duty cycle of the RO rather than the period. However, this method reduces the effective utilization of process variations, which limits its implementation in mature advanced manufacturing processes. In this study, a splitting duty-cycle (SDC) PUF was proposed to balance the effective extraction of process variations and robustness under voltage and temperature changes. The sensibility formula between the performance of SDC PUF and process, voltage, and temperature was established through a circuit model and statistical methodology, and the comprehensive characteristics of SDC PUF were analyzed theoretically. Next, 16 SDC PUFs with 128-bit responses were implemented and measured on a Xilinx Virtex-7 device. The experimental results revealed that the average native reliability of SDC PUF was 98.97%, and the reliability was 97.32% under various voltage and temperature conditions. This result revealed advantages over the DC PUF implemented in the same device. The uniqueness of the SDC PUF was 50.42%, and it passed the NIST SP 800-22 randomness and autocorrelation function tests.

True random number generators (TRNGs), as an important component of security systems, have received a lot of attention for their related research. The previous researches have provided a large number of TRNG solutions, however, they still failed to reach an excellent tradeoff in various performance metrics. This article presents a shift-registers metastability-based TRNG, which is implemented by compact reference units and comparison units. By forcing the D flip-flops in the shift-registers into the metastable state, it optimizes the problem that the conventional metastability entropy sources consume excessive hardware resources. And a new method of metastable randomness extraction is used to reduce the bias of metastable output. The proposed TRNG is implemented in Xilinx Spartan-6 and Virtex-6 FPGAs, which generate random sequences that pass the NIST SP800-22, NIST SP800-90B tests and show excellent robustness to voltage and temperature variations. This TRNG can consume only 3 slices of the FPGA, but it has a high throughput rate of 25 Mbit/s. In comparison with state-of-the-art FPGA-compatible TRNGs, the proposed TRNG achieves the highest figure of merit FOM, which means that the proposed TRNG significantly outperforms previous researches in terms of hardware resources, throughput rate, and operating frequency tradeoffs.

The entropy source structure with embedded XOR gates in a ring oscillator (RO) as a true random number generator (TRNG) can improve the speed of accumulating jitter in the oscillator. However, the XOR gate has a certain response time to the input change, and when the input changes too fast, the XOR gate will output short pulses. In this paper, we propose a TRNG design based on a multi-ring convergence oscillator (MRCO) making use of the characteristics of short pulses. We study the output of the XOR gate when facing different inputs. By modeling the time of a fibonacci ring oscillator (FIRO) as an example, we find that the loss of short pulses in an inverter chain is the reason for making the FIRO enter into periodic oscillation. This phenomenon suppresses the accumulation of jitter and occurs periodically in existing structures. Our proposed structure uses independent sub-rings to accumulate jitter, allowing the main-ring to quickly generate short pulses to provide analog randomness. The proposed TRNG design is implemented in Xilinx Virtex-6 FPGA. The experimental results show that it has the highest ratio of throughput rate to hardware resources. The generated random sequence pass both NIST SP800-22 test and NIST SP800-90B test.

Since the physical defects such as resistive open and leakage in through silicon vias (TSVs) caused by immature manufacturing techniques tend to undermine the reliability and yield of 3-D integrated circuits, it is very important to test the TSV as early as possible in the fabrication process. There are some shortcomings in the existing prebond TSV test techniques, such as incomprehensive fault coverage, large area overhead, and additional test time. To overcome these problems, a noninvasive solution for prebond TSV test based on pulse shrinking is proposed in this paper. This method makes use of the fact that defects in TSV lead to variation in the propagation delay-the rise and fall times are first transformed into pulse width, and the pulse shrinking technique is used to digitize the pulse width into a digital code which is then compared with an expected value for a fault-free TSV. Experiments on defect detection are carried out using HSPICE simulations with realistic models for 45-nm CMOS technology. The results show that the proposed method performs better than the existing methods in terms of fault coverage, area overhead, and test time.