Files in this item



application/pdfKONIGSMARK-DISSERTATION-2017.pdf (3MB)Restricted to U of Illinois
(no description provided)PDF


Title:Hardware security design from circuits to systems
Author(s):Konigsmark, Sven Tenzing Choden
Director of Research:Chen, Deming; Wong, Martin D F
Doctoral Committee Chair(s):Chen, Deming; Wong, Martin D F
Doctoral Committee Member(s):Sanders, William H; Nicol, David M
Department / Program:Electrical & Computer Eng
Discipline:Electrical & Computer Engr
Degree Granting Institution:University of Illinois at Urbana-Champaign
Subject(s):Hardware Security
Physically Unclonable Function
Hardware Trojan horse
Invasive Attacks
Carbon Nanotubes
Abstract:The security of hardware implementations is of considerable importance, as even the most secure and carefully analyzed algorithms and protocols can be vulnerable in their hardware realization. For instance, numerous successful attacks have been presented against the Advanced Encryption Standard, which is approved for top secret information by the National Security Agency. There are numerous challenges for hardware security, ranging from critical power and resource constraints in sensor networks to scalability and automation for large Internet of Things (IoT) applications. The physically unclonable function (PUF) is a promising building block for hardware security, as it exposes a device-unique challenge-response behavior which depends on process variations in fabrication. It can be used in a variety of applications including random number generation, authentication, fingerprinting, and encryption. The primary concerns for PUF are reliability in presence of environmental variations, area and power overhead, and process-dependent randomness of the challenge-response behavior. Carbon nanotube field-effect transistors (CNFETs) have been shown to have excellent electrical and unique physical characteristics. They are a promising candidate to replace silicon transistors in future very large scale integration (VLSI) designs. We present the Carbon Nanotube PUF (CNPUF), which is the first PUF design that takes advantage of unique CNFET characteristics. CNPUF achieves higher reliability against environmental variations and increases the resistance against modeling attacks. Furthermore, CNPUF has a considerable power and energy reduction in comparison to previous ultra-low power PUF designs of 89.6% and 98%, respectively. Moreover, CNPUF allows a power-security tradeoff in an extended design, which can greatly increase the resilience against modeling attacks. Despite increasing focus on defenses against physical attacks, consistent security oriented design of embedded systems remains a challenge, as most formalizations and security models are concerned with isolated physical components or a high-level concept. Therefore, we build on existing work on hardware security and provide four contributions to system-oriented physical defense: (i) A system-level security model to overcome the chasm between secure components and requirements of high-level protocols; this enables synergy between component-oriented security formalizations and theoretically proven protocols. (ii) An analysis of current practices in PUF protocols using the proposed system-level security model; we identify significant issues and expose assumptions that require costly security techniques. (iii) A System-of-PUF (SoP) that utilizes the large PUF design-space to achieve security requirements with minimal resource utilization; SoP requires 64% less gate-equivalent units than recently published schemes. (iv) A multilevel authentication protocol based on SoP which is validated using our system-level security model and which overcomes current vulnerabilities. Furthermore, this protocol offers breach recognition and recovery. Unpredictability and reliability are core requirements of PUFs: unpredictability implies that an adversary cannot sufficiently predict future responses from previous observations. Reliability is important as it increases the reproducibility of PUF responses and hence allows validation of expected responses. However, advanced machine-learning algorithms have been shown to be a significant threat to the practical validity of PUFs, as they can accurately model PUF behavior. The most effective technique was shown to be the XOR-based combination of multiple PUFs, but as this approach drastically reduces reliability, it does not scale well against software-based machine-learning attacks. We analyze threats to PUF security and propose PolyPUF, a scalable and secure architecture to introduce polymorphic PUF behavior. This architecture significantly increases model-building resistivity while maintaining reliability. An extensive experimental evaluation and comparison demonstrate that the PolyPUF architecture can secure various PUF configurations and is the only evaluated approach to withstand highly complex neural network machine-learning attacks. Furthermore, we show that PolyPUF consumes less energy and has less implementation overhead in comparison to lightweight reference architectures. Emerging technologies such as the Internet of Things (IoT) heavily rely on hardware security for data and privacy protection. The outsourcing of integrated circuit (IC) fabrication introduces diverse threat vectors with different characteristics, such that the security of each device has unique focal points. Hardware Trojan horses (HTH) are a significant threat for IoT devices as they process security critical information with limited resources. HTH for information leakage are particularly difficult to detect as they have minimal footprint. Moreover, constantly increasing integration complexity requires automatic synthesis to maintain the pace of innovation. We introduce the first high-level synthesis (HLS) flow that produces a threat-targeted and security enhanced hardware design to prevent HTH injection by a malicious foundry. Through analysis of entropy loss and criticality decay, the presented algorithms implement highly resource-efficient targeted information dispersion. An obfuscation flow is introduced to camouflage the effects of dispersion and reduce the effectiveness of reverse engineering. A new metric for the combined security of the device is proposed, and dispersion and obfuscation are co-optimized to target user-supplied threat parameters under resource constraints. The flow is evaluated on existing HLS benchmarks and a new IoT-specific benchmark, and shows significant resource savings as well as adaptability. The IoT and cloud computing rely on strong confidence in security of confidential or highly privacy sensitive data. As (differential) power attacks can take advantage of side-channel leakage to expose device-internal secrets, side-channel leakage is a major concern with ongoing research focus. However, countermeasures typically require expert-level security knowledge for efficient application, which limits adaptation in the highly competitive and time-constrained IoT field. We address this need by presenting the first HLS flow with primary focus on side-channel leakage reduction. Minimal security annotation to the high-level C-code is sufficient to perform automatic analysis of security critical operations with corresponding insertion of countermeasures. Additionally, imbalanced branches are detected and corrected. For practicality, the flow can meet both resource and information leakage constraints. The presented flow is extensively evaluated on established HLS benchmarks and a general IoT benchmark. Under identical resource constraints, leakage is reduced between 32% and 72% compared to the baseline. Under leakage target, the constraints are achieved with 31% to 81% less resource overhead.
Issue Date:2017-03-13
Rights Information:Copyright 2017 Choden Konigsmark
Date Available in IDEALS:2017-08-10
Date Deposited:2017-05

This item appears in the following Collection(s)

Item Statistics