Hardware Security and Tamper Protection

I. Introduction to Integrated Circuit Physical Security and Threat Landscape

A. Context and Strategic Importance of Hardware Hardening

The security landscape for integrated circuits (ICs) has shifted dramatically, necessitating robust hardware defenses as attackers increasingly push low into the platform stack, moving below conventional software controls.1 This physical hardening is strategically vital because the physical platform represents the immutable first layer in any layered security approach. It provides the initial protections required to ensure that all higher-layer security controls—such as operating systems, secure firmware, and applications—can be fundamentally trusted.2 Without this fortified foundation, the integrity of the entire system, including advanced concepts like Confidential Computing or Secure Boot, is potentially compromised.

The necessity of IC hardening addresses both known vulnerabilities and persistent threat models, notably the Hardware Trojan (HT). HTs are malicious, intentional modifications inserted into an electronic circuit designed to cause incorrect behavior, leak sensitive information, or launch Denial of Service (DoS) attacks.4 These modifications can be subtle, such as Trojans embedded in the clock grid capable of changing clock frequency, inserting glitches, or increasing signal skew to specific functional modules, thereby launching targeted fault attacks.5 Hardening efforts are not merely reactive; they are foundational, providing the necessary assurance that the underlying platform itself is reliable and untampered, validating the subsequent secure operations of the system.

B. Classification of Physical Attacks on ICs: Fault Injection vs. Invasive Tampering

Physical attacks on ICs are typically categorized into two primary methodologies based on their invasiveness and objective: Fault Injection Attacks (FIA) and Invasive Tampering.

Fault Injection Attacks (FIA) are methods, often non-invasive or semi-invasive, aimed at introducing transient errors to manipulate the device’s logic state, forcing it to skip instructions, bypass security checks, or corrupt data.6 Key types include:

• Voltage Glitching: This involves manipulating the power supply to create brief, targeted disruptions in voltage delivery.7

• Clock Glitching: This method injects timing faults into the device’s clock signals, disrupting the standard data processing timeline.7

• Electromagnetic Fault Injection (EMFI) and Laser Fault Injection (LFI): These sophisticated techniques use external energy sources—electromagnetic pulses or ionizing radiation—to directly flip individual bits or alter the state of logic gates, leading to predictable errors in the output.6

Invasive Tampering, conversely, involves direct physical access to the chip to modify functionality or extract data. Examples include micro-probing sensitive wires to read or force an equipotential, cutting wires, re-routing critical signal paths, or altering onboard memory such as ROM or fuses.10

C. Regulatory and Certification Drivers

The necessity of robust hardware security is supported by international regulatory guidance. The National Institute of Standards and Technology (NIST) has issued detailed frameworks, such as NIST IR 8320, which focuses on hardware-enabled security, emphasizing its role in enabling a layered security approach for cloud data centers and edge computing.2 These guidelines underscore the need for rigorous platform integrity verification and define the roles of specialized components such as Hardware Security Modules (HSMs) and Trusted Platform Modules (TPMs).1 NIST also publishes specific guidelines covering fundamental components like BIOS protection and integrity measurement, recognizing the low-level vulnerabilities that must be addressed.12 Furthermore, commercial hardware security modules are validated and certified to stringent, globally recognized standards, including FIPS 140-2 and Common Criteria, which provide objective assurance of their tamper-resistant capabilities and resilience.13

II. Advanced Analysis of Clock Fault Injection Mechanisms and Countermeasures

A. Detailed Taxonomy of Clock Attack Vectors

Clock fault injection is a powerful attack vector because it targets the fundamental timing integrity of the IC. The primary goal of clock glitching is to momentarily violate the setup or hold time requirements of the device's flip-flops. If timed precisely, this violation can cause the digital logic to misexecute or skip critical instructions, compromising security.7

Attackers use various methods to generate these glitches, often assuming they have access to the external clock signal.14 These generation methods are broadly categorized into two families:

1. Combine Shifted Clocks (CSC): This approach uses two clocks that are slightly phase-shifted relative to the nominal clock signal.

2. Combine Different Clock Frequencies (CDCF): This method involves temporarily mixing or changing clock frequencies to generate the desired disruption.14

Beyond external injection, internal threats are also critical. Hardware Trojans can be strategically inserted into the clock grid to launch highly targeted fault attacks. These Trojans can maliciously change the clock's frequency, insert precise glitches, or increase signal skew specifically impacting vulnerable functional modules of the chip.5

B. Effects of CFI on Digital Logic and Cryptographic Primitives

The immediate, transactional effect of a successful clock glitch is the failure of a logic path, leading to instruction skipping or corruption. This vulnerability is routinely exploited to bypass authentication sequences, allowing the attacker to skip critical security checks.7

However, the impact extends beyond cryptographic key extraction. Recent studies demonstrate that clock glitch fault attacks can induce serious misclassifications in Deep Neural Networks (DNNs) deployed on embedded edge devices.15 For instance, an attacker could force a DNN designed for image recognition to misidentify data with high probability. This indicates that hardware fault attacks pose an existential threat not only to cryptographic assets but also to the functional integrity of emerging AI applications, such such as those used in autonomous systems.15

Designing circuits that are resistant to such manipulation requires complex fault-resistance models. Verification of security must account for stringent parameters, including the maximum number of fault events per clock cycle, the total number of clock cycles faults can occur within, and the precise set of allowed fault types.16 Rigorously verifying this fault-resistance, especially at the gate level, is a complex, computationally intensive challenge.

C. Clock Integrity Monitoring (CIM) Hardware Architecture

To counteract temporal attacks, advanced architectural approaches utilize dedicated on-chip security Intellectual Property (IP). A highly efficient strategy involves designing components that convert diverse physical threats into a single, measurable metric: timing stress.17

This unified detection mechanism is predicated on the physical reality that multiple environmental or injection attacks ultimately impact the critical path delay margin of the digital logic. For example:

1. Overclocking directly reduces the clock period, causing a critical path violation.

2. Reducing the supply voltage (a power glitch) increases the propagation time of combinational logic.

3. Increasing the temperature also increases the logic propagation time.17

In each case, the effect is measured as a timing stress. This centralization of detection means that dedicated digital sensor IP can monitor these stresses in real-time. This IP core is fully digital, designed to be lightweight, and can be subtly blended into the rest of the chip design to make it difficult for an attacker to identify or circumvent.17 Upon detection of a timing stress exceeding the acceptable threshold, the sensor IP immediately triggers a real-time hardware alarm.17 Furthermore, monitoring circuitry utilizing Time-to-Digital Converters (TDCs) integrated alongside Ring Oscillators is employed to monitor precise timing stability and subtle changes in power consumption patterns indicative of an attack.18 This architectural strategy ensures that a single, verifiable IP core can address four major fault attack families, significantly simplifying the overall design for security (DfS).

The relationship between various fault types and the resulting timing stress is summarized in the table below:

Table 1: Integrated Fault Injection Detection via Timing Stress Conversion

Fault Injection Stress	Physical Effect on Circuit	Manifestation in Digital Sensor IP	Citing Snippets
Input Clock Frequency (Glitch/Overclock)	Reduction of clock period/Timing violation	Measured as direct timing stress (period change)	17
Input Voltage (Glitch/Underfeeding)	Increase in logic propagation delay	Measured as timing stress (critical path delay)	17
Temperature (Heating)	Increase in logic propagation delay	Measured as timing stress (critical path delay)	17
Radiation (Laser/EMFI)	Bit set/reset in registers/Transient delay	Measured as localized timing stress or state error	6

D. Integrating Resistance to Side-Channel Attacks (SCA)

Side-channel attacks (SCA), such as Differential Power Analysis (DPA), Simple Power Analysis (SPA), and Correlation Electromagnetic Analysis (CEMA), are inexpensive and exploit the physical information leaked by electronic systems.19 This leakage is intrinsically linked to the time-dependent, clocked operations of the digital circuit. DPA, in particular, statistically analyzes multiple power consumption traces from cryptographic operations to extract secret information, such as keys used in algorithms like AES or DES.9

An effective defense architecture requires layered countermeasures implemented via hardware (DPA resistant cores), software (DPA resistant libraries), or a combination of both.19 Hardware-layer DPA countermeasures involve architectural changes designed to mitigate leakage, introduce randomness, balance hardware and software operations, and incorporate amplitude and temporal noise.20

It is important to recognize that while hardware monitoring detects the stress, highly complex applications like AI/ML often require supplementary defenses. For DNNs, for instance, specialized algorithmic countermeasures applied to critical functions, such as Softmax and Sigmoid, can effectively prevent misclassification incidents even if a fault event is detected.15 This illustrates the need for a hybrid fault tolerance strategy where hardware monitors provide general environmental security, and specialized algorithmic safeguards ensure run-time resilience for application-specific critical data.

III. Securing the Reset Mechanism Against Malicious Injection

A. The Vulnerability of the Reset Signal in IC State Transition

The reset signal is one of the most critical control vectors in an IC, responsible for fundamental state initialization and recovery. It ensures the microcontroller starts executing code from the correct memory location during initial power-up (Power-On Reset, POR) and provides necessary mechanisms for recovering from runtime instability (Brown-Out Reset, BOR, or Watchdog Timer Reset).22

Because the reset signal controls the most fundamental state transition, it is a prime target for malicious injection attacks. These attacks seek to manipulate system state, resulting in a Denial of Service (DoS) or forcing the device into an insecure or transiently vulnerable state, often by disrupting the expected orderly power-up sequence.23 Furthermore, reset injection often overlaps causally with other fault injection techniques. Voltage glitches or Electromagnetic Fault Injection (EMFI) can be used to momentarily disrupt supply voltage integrity, triggering or mimicking a reset condition.6

B. Secure Reset Circuit Implementation

Secure IC design mandates robust implementations of foundational reset circuits:

• Power-On Reset (POR): This ensures the system begins in a known and safe state, managing the necessary orderly and predictable sequence of events required for components (like bias current supplies) to stabilize during power-up.24

• Brown-Out Reset (BOR): This circuit detects conditions where the supply voltage drops below an operational threshold (a brown-out). BOR is vital as it prevents the erratic or undefined behavior that low voltage causes by forcing a secure reset.22

To harden these components against malicious fault injection, on-chip Supply Voltage Supervisor (SVS) circuitry must be utilized to continuously monitor for overvoltage or undervoltage conditions.25 Robust BOR implementation is paramount for physical security, as it acts as a primary, non-clock-based defense against generalized voltage fault injection. An attacker attempting to inject a voltage glitch to skip instructions must precisely time the glitch to fall below the level needed for correct logic operation but remain above the threshold that would trigger the secure, built-in BOR/SVS mechanisms.25 Hardening the BOR circuit (by employing robust hysteresis and tight voltage thresholds) forces the attacker into a full system reset rather than instruction skipping, defeating the transient nature of the glitch attack.

Advanced Reset Synchronization for Metastability Mitigation

Beyond environmental protection, internal digital integrity is maintained through careful reset synchronization. The complexity of reset design often introduces issues, notably metastability problems, when asynchronous reset signals are improperly handled, leading to unpredictable internal logic states.26 The industry standard solution involves using a reset synchronizer circuit. This synchronizer allows for the rapid, asynchronous application of a reset signal but ensures its removal is synchronous to the system clock. This prevents metastability and guarantees the safe, predictable restoration of normal functional operation.26

The comparison of key reset mechanisms and their hardening requirements is summarized below:

Table 2: Comparison of Secure Initialization and Reset Integrity Requirements

Mechanism	Trigger	Security Imperative	Hardening Countermeasure
Power-On Reset (POR)	Initial power application	Forces known, safe starting state, sequenced startup	Stable voltage monitoring, proper sequencing logic
Brown-Out Reset (BOR)	Supply voltage dips below threshold	Prevents erratic behavior/instruction misexecution	Hysteresis, robust noise filtering, tight thresholds 22
Malicious Reset (FI)	External EMFI or intentional signal injection	Prevents unintended state transition (DoS/bypass)	Tamper detection circuitry, monitoring reset reason 6
Reset Synchronizer	Asynchronous reset removal	Ensures synchronous release to prevent digital metastability	Multi-stage flip-flop implementation for safe assertion removal 26

C. Secure Initialization and Post-Reset Integrity Verification

Following any reset event, whether POR or malicious injection, the system must verify its integrity before resuming normal operation. The device should check the reason for the reset and take appropriate application-level recovery actions.25 Crucially, the secure initialization process must transition immediately into a hardware-backed Secure Boot sequence. Secure Boot verifies the integrity of the initial firmware before any execution is allowed.28 If integrity checks fail (suggesting tampering or improper initialization), the device must be driven into a protected or locked state.28 Furthermore, tools like QEMU based fault injection test tools can be integrated with continuous integration pipelines to test hardening code against fault models, such as the common instruction skip fault, verifying that the embedded defensive labels trigger correctly upon tampering.29

IV. Comprehensive Physical Tamper Detection and Response

Physical security requires a defense-in-depth strategy, integrating passive resistance, continuous active detection, and immediate, irreversible response mechanisms.

A. Classification of Physical Tamper Threats

Physical tamper threats are designed to gain unauthorized access to data or control. Invasive techniques include micro-probing to read or force equipotential lines, wire cutting, and actively altering ROM or fuse settings.10 Semi-invasive techniques rely on localized energy sources such as laser beams, light spots, or electromagnetic fields, which can cause transient faults like bit set or reset events in registers.6

B. Active Shielding and Tamper Meshes (The Active Defense Layer)

Passive protection methods like potting and tamper-proof coatings 11 are often insufficient against determined adversaries. Therefore, continuous active monitoring through anti-tamper circuitry is essential.30

Conductive Mesh Design

Modern chips incorporate conductive meshes, often referred to as Active Shield IP, placed over sensitive circuit areas to protect core components like metal routing and transistors.10 These meshes are conductive grids embedded on or within the IC layers.11 For sophisticated integrated circuits, the mesh can be digitally controlled and interleaved with the Power/Ground (P/G) network in the topmost layers, avoiding the need to sacrifice an entire metal layer solely for security.10

Monitoring and Detection

Active tamper detection relies on monitoring the integrity of this mesh. A state-of-the-art technique involves continuously outputting a randomized, cryptographically generated pattern on an output pin (TAMP_OUT) which is externally shorted to an input pin (TAMP_IN).10 If a physical intrusion—such as cutting, shorting, or micro-probing—occurs, the signal comparison fails instantly, triggering a real-time hardware alarm.33 Using randomized patterns and potentially AC signals rather than simple DC resistance monitoring provides greater resistance against sophisticated probing attempts.10 Future systems may incorporate advanced methods like Anti-Tamper Radio (ATR), which analyzes radio wave propagation within the device enclosure to detect unauthorized alterations, potentially enhanced by Reconfigurable Intelligent Surfaces (RIS) to improve robustness and sensitivity.34

C. Environmental and Localized Fault Detection

Integrated monitoring systems provide crucial complementary defenses. PVT (Process, Voltage, Temperature) monitors, traditionally used for silicon lifecycle management and performance optimization, are vital security components.35 These systems detect unauthorized, rapid, or sustained environmental changes outside nominal operating ranges, linking environmental manipulation (such as heating or cooling) to potential fault injection attempts.17

These PVT monitors work in conjunction with the Digital Sensor IP described earlier. For instance, a rise in temperature might be detected by a PVT monitor, but the system distinguishes between a non-malicious environmental fluctuation and a security-critical failure by simultaneously verifying whether the temperature increase has caused the logic propagation delay to exceed the safe timing margin.17 This synergy ensures highly reliable security alarms with reduced false-positive rates, differentiating environmental stress from security-critical timing failures. Furthermore, dedicated digital sensors are employed to detect specific directed energy attacks, such as laser or electromagnetic radiation, by measuring the resulting timing stress or bit flips.6

D. Tamper Response Strategies

Since absolute tamper prevention is physically impossible, a robust system must focus on mitigating the attack outcome by maximizing the cost and minimizing the reward.36 The primary mechanism for achieving this is the immediate and irreversible response upon detection.

Robust response mechanisms include:

1. Device Blocking: Temporarily or permanently rendering the device inoperable.31

2. Memory Zeroization: This is the most critical response, involving the immediate erasure of Critical Security Parameters (CSPs), such as cryptographic keys and sensitive key data, thereby neutralizing the attacker’s objective.31

3. Physical Destruction: In high-security applications, the device may be physically destroyed, often by significantly raising the supply voltage using an onboard boost converter to burn out components. Other methods include using fragile glass substrates that shatter upon detection of heat or vibration changes.31

These responses, particularly zeroization, ensure that even if an attacker successfully breaches the physical perimeter, they cannot extract exploitable data, rendering the expensive intrusion futile. Additionally, techniques are often implemented to ensure the device retains evidence of tampering (tamper evidence) for subsequent forensic analysis.31

V. Architectural Frameworks for Hardened Silicon

A. Establishing the Hardware Root of Trust (HRoT)

A trusted computing architecture must be anchored by an immutable Hardware Root of Trust (HRoT), which serves as the physical component responsible for establishing secure identity, generating keys, and validating the integrity of the initial boot code.2

Physical Unclonable Functions (PUFs)

The foundation of HRoT often rests on Physical Unclonable Functions (PUFs), such as NeoPUF technology, which leverage the unique, uncontrollable manufacturing variations of the silicon to generate a unique and unclonable identifier (UID) and root key directly within the chip.37 The PUF Root of Trust (PUFrt) component facilitates this generation process and is equipped with a resilient anti-tamper shell to protect the key generation circuitry itself.37 The resulting hardware root key is never allowed to leave the chip boundary.37

Tamperproof Secure Storage

Complementing the PUF is the requirement for tamperproof storage of sensitive data, configuration settings, and validated boot code. Secure OTP (One-Time Programmable) memory, typically implemented using anti-fuse technology, provides secure storage for keys, data, and secrets in use, transit, or at rest.37 Crucially, true HRoT is implemented as a system where both the key generation block (PUFrt) and the secure data storage block (Secure OTP) incorporate their own independent, resilient anti-tamper shells to provide comprehensive protection against hardware attacks.37 This interdependency ensures that the integrity of both the unclonable identity and the securely stored state data is maintained, confirming a holistic design methodology.

The essential components of hardware tamper defense and the HRoT framework are summarized below:

Table 3: Hardening Techniques for Physical Tamper Defense

Technique Type	Example Implementation	Security Function/Target Threat	Response Mechanism
Active Shielding (IC Level)	Conductive Mesh/Active Shield IP	Detects wire cutting, micro-probing, routing alteration 10	Real-time alarm, Zeroization 31
Tamperproof Storage	Secure OTP (Anti-fuse memory)	Secure storage of CSPs (keys, boot hashes) 37	Dedicated anti-tamper shell, Data protection in transit/rest
Root Key Generation	PUFrt (NeoPUF technology)	Generates unique, unclonable device identity/root key 37	Integrated anti-tamper shell, Key never leaves the chip
Response Systems	High voltage boost converter, key erasure circuit	Defeats data extraction (physical destruction, zeroization) 31	Immediate, irreversible, high-speed reaction

B. Utilizing Certified Security Modules: HSMs and TPMs

Dedicated security modules are integral to maintaining trust within the HRoT framework. Hardware Security Modules (HSMs) are specialized, hardened, tamper-resistant computing devices specifically designed to safeguard and manage cryptographic keys, ensuring key generation, protection, and management are secured from compromise.1 These modules are validated against rigorous certification standards, such as FIPS 140-2 Level 3.13

The Trusted Platform Module (TPM) is a specialized form of HSM that provides fundamental root of trust functions, including the ability to generate cryptographic keys and protect small amounts of sensitive information, such as cryptographic hash measurements and passwords, essential for maintaining platform integrity.1

C. Integration Strategy: Layered Security and Defense-in-Depth

The ultimate goal of chip hardening is to ensure a reliable Chain of Trust (CoT). This CoT originates at the hardware layer, anchored by the HRoT, and transitively validates subsequent layers of firmware and software.1 This requires that cryptographic primitives, even lightweight versions suitable for resource-constrained devices like IoT, are securely integrated into the boot process for firmware integrity verification.28

A crucial component of the strategy must be rigorous verification. Beyond implementation, rigorous verification of fault-resistance is critical, especially at the gate level, which requires specialized tools to analyze resilience against varying fault types and cycles.16 The combined strategy of unified timing stress detection, secure reset management, active physical shielding, and HRoT implementation creates a highly resilient system capable of surviving sophisticated temporal and physical attacks.

VI. Conclusion and Future Resilience Recommendations

A. Synthesis of State-of-the-Art Hardening Techniques

Modern chip hardening against fault injection and tampering is characterized by the convergence of temporal and physical defenses, moving beyond simple protection to active, intelligent detection and immediate, irreversible response. Temporal integrity is maintained through sophisticated techniques like the Digital Sensor IP, which standardizes diverse fault injection stresses (clock, voltage, temperature, radiation) into a measurable timing failure.17 State management is secured through robust Power-On Reset (POR) and Brown-Out Reset (BOR) circuits, complemented by asynchronous reset synchronizers that guarantee safe state transitions.22

The physical perimeter is protected by active shielding, utilizing randomized conductive meshes that continuously monitor for intrusions and trigger immediate alarms upon disruption.10 This entire defensive architecture is anchored by a verifiable Hardware Root of Trust (HRoT), built on Physical Unclonable Function (PUF) technology for identity and anti-fuse Secure OTP for tamperproof storage.37

B. Recommendations for Future Resilience

As the attack surface continues to evolve, future hardware resilience strategies must focus on enhancing verification and anticipating emerging threats:

• Advanced Fault Verification: Current methods for rigorously proving fault-resistance, especially gate-level analysis, remain computationally challenging. Continued investment in efficient, complete verification methodologies is necessary to ensure resilience against increasingly sophisticated fault models.16

• Integration of AI/ML: While AI and Deep Neural Networks (DNNs) are themselves targets of fault injection 15, AI also represents a future trend in both attack (designing zero-overhead, intelligent hardware Trojans) and defense (developing highly sensitive detection techniques).38 Security architectures must incorporate AI-driven analysis for intrusion detection.

• Quantum Readiness: Given the emergence of quantum computing threats, hardening solutions must transition to be Post-Quantum Cryptography (PQC)-ready, particularly within the HRoT framework, to secure key generation and storage against future cryptanalytic threats.37

Works cited

1. NISTIR 8320, 8320A, 8320B, Hardware-Enabled Security, accessed October 7, 2025, https://csrc.nist.gov/csrc/media/Presentations/2023/hardware-enabled-security/images-media/Jan-24-2023-bartock.pdf

2. NIST IR 8320, accessed October 7, 2025, https://nvlpubs.nist.gov/nistpubs/ir/2022/Nist.IR.8320.pdf

3. IR 8320, Hardware-Enabled Security: Enabling a Layered Approach to Platform Security for Cloud and Edge Computing Use Cases | CSRC, accessed October 7, 2025, https://csrc.nist.gov/pubs/ir/8320/final

4. Hardware Trojans - Prevention, Detection, Countermeasures (A Literature Review) - DTIC, accessed October 7, 2025, https://apps.dtic.mil/sti/tr/pdf/ADA547668.pdf

5. TRUSTWORTHY HARDWARE: IDENTIFYING AND CLASSIFYING HARDWARE TROJANS - Google Research, accessed October 7, 2025, https://research.google.com/pubs/archive/37398.pdf

6. SoK: A Beginner-Friendly Introduction to Fault Injection Attacks - arXiv, accessed October 7, 2025, https://arxiv.org/html/2509.18341v1

7. What Is a Glitching Attack in Cybersecurity and Why Does It Matter? - Huntress, accessed October 7, 2025, https://www.huntress.com/cybersecurity-101/topic/what-is-glitching-cybersecurity-attack

8. FaultDetective - CHES - IACR, accessed October 7, 2025, https://ches.iacr.org/2024/papers-issue-4/4_92.pdf

9. Modern Hardware Security: A Review of Attacks and Countermeasures - arXiv, accessed October 7, 2025, https://arxiv.org/html/2501.04394v1

10. Active shield IP to protect integrated circuits - Secure-IC, accessed October 7, 2025, https://www.secure-ic.com/products/securyzr/security-ip/anti-tamper/active-shield-ip/

11. What is Tamper-Proofing in PCBs? - PCB Directory, accessed October 7, 2025, https://www.pcbdirectory.com/community/what-is-tamper-proofing-in-pcbs

12. Hardware Security | CSRC, accessed October 7, 2025, https://csrc.nist.gov/projects/hardware-security

13. What is a Hardware Security Module (HSM) & its Services? - Entrust, accessed October 7, 2025, https://www.entrust.com/resources/learn/what-are-hardware-security-modules

14. A Review on Evaluation and Configuration of Fault Injection Attack Instruments to Design Attack Resistant MCU-Based IoT Applications - MDPI, accessed October 7, 2025, https://www.mdpi.com/2079-9292/9/7/1153

15. Clock Glitch Fault Attacks on Deep Neural Networks and Their Countermeasures - PMC, accessed October 7, 2025, https://pmc.ncbi.nlm.nih.gov/articles/PMC12074318/

16. SAT-based Formal Verification of Fault Injection Countermeasures for Cryptographic Circuits - CHES, accessed October 7, 2025, https://ches.iacr.org/2024/papers-issue-4/4_7.pdf

17. Digital Sensor IP - Secure-IC, accessed October 7, 2025, https://www.secure-ic.com/products/securyzr/security-ip/anti-tamper/digital-sensor/

18. Investigating Time-Digital-Converters for Hardware Security in FPGAs - Oregon State University, accessed October 7, 2025, https://ir.library.oregonstate.edu/downloads/nc580w33z?locale=en

19. DPA Countermeasures - DPA Security Solutions - Rambus, accessed October 7, 2025, https://www.rambus.com/security/dpa-countermeasures/

20. SCA Resistant Cores with DPA and FIA - Rambus, accessed October 7, 2025, https://www.rambus.com/security/dpa-countermeasures/sca-resistant-cores/

21. Modern Hardware Security: A Review of Attacks and Countermeasures - arXiv, accessed October 7, 2025, https://arxiv.org/pdf/2501.04394

22. How to Design Reliable Reset Circuits for Embedded Microcontrollers, accessed October 7, 2025, https://runtimerec.com/how-to-design-reliable-reset-circuits-for-embedded-microcontrollers/

23. ICS Techniques - MITRE ATT&CK®, accessed October 7, 2025, https://attack.mitre.org/techniques/ics/

24. Self-Starting Can Be Hard to Do - Understanding Power-on-Reset Requirements in Regulators and Bandgaps | Analog Devices, accessed October 7, 2025, https://www.analog.com/en/resources/technical-articles/selfstarting-can-be-hard-to-do--understanding-poweronreset-requirements-in-regulators-and-bandgaps.html

25. System-Level Tamper Protection Using MSP MCUs - Texas Instruments, accessed October 7, 2025, https://www.ti.com/lit/pdf/slaa715

26. Asynchronous & Synchronous Reset Design Techniques - Part Deux - LCDM-ENG, accessed October 7, 2025, https://lcdm-eng.com/papers/snug03_reset_part_deux.pdf

27. Asynchronous & Synchronous Reset Design Techniques - Part Deux, accessed October 7, 2025, https://docenti.ing.unipi.it/f.baronti/didattica/SE/2015/Ppt/CummingsSNUG2003Boston_Resets.pdf

28. A lightweight Secure Boot Mechanism for Protecting the Firmware of IOT Devices⋆ - CEUR-WS.org, accessed October 7, 2025, https://ceur-ws.org/Vol-4048/paper19.pdf

29. HW Fault Injection Mitigation - Trusted Firmware, accessed October 7, 2025, https://www.trustedfirmware.org/docs/TF-M_fault_injection_mitigation.pdf

30. Hardware Security | KTH, accessed October 7, 2025, https://www.kth.se/social/files/59102ef5f276540f03507109/hardware_security__2017_05_08.pdf

31. Hardware-Based Methods for Electronic Device Protection against Invasive and Non-Invasive Attacks - MDPI, accessed October 7, 2025, https://www.mdpi.com/2079-9292/12/21/4507

32. What Makes A Chip Tamper-Proof? - Semiconductor Engineering, accessed October 7, 2025, https://semiengineering.com/what-makes-a-chip-tamper-proof/

33. Hello, and welcome to this presentation of the STM32 Tamper and backup registers. It covers the main features of this peripheral, accessed October 7, 2025, https://www.st.com/resource/en/product_training/STM32MP1-Security-Tamper_TAMP.pdf

34. Anti-Tamper Radio meets Reconfigurable Intelligent Surface for System-Level Tamper Detection - arXiv, accessed October 7, 2025, https://arxiv.org/html/2503.14279v1

35. Synopsys SLM Process, Voltage & Temperature Monitor IP, accessed October 7, 2025, https://www.synopsys.com/solutions/silicon-lifecycle-management/environmental-monitors/slm-pvt-monitor-ip.html

36. Tamperproofing - Wikipedia, accessed October 7, 2025, https://en.wikipedia.org/wiki/Tamperproofing

37. Products | PUF-based Security IP Solutions | PUFsecurity, accessed October 7, 2025, https://www.pufsecurity.com/products/

38. Hardware Trojans in Chips: A Survey for Detection and Prevention - MDPI, accessed October 7, 2025, https://www.mdpi.com/1424-8220/20/18/5165