First Advisor

John M. Acken

Term of Graduation

Summer 2023

Date of Publication

7-18-2023

Document Type

Dissertation

Degree Name

Doctor of Philosophy (Ph.D.) in Electrical and Computer Engineering

Department

Electrical and Computer Engineering

Language

English

Subjects

Data encryption (Computer science)

DOI

10.15760/etd.3651

Physical Description

1 online resource (xv, 171 pages)

Abstract

Power Side Channel Attacks have continued to be a major threat to cryptographic devices. Hence, it will be useful for designers of cryptographic systems to systematically identify which type of power Side Channel Attacks their designs remain vulnerable to after implementation. It’s also useful to determine which additional vulnerabilities they have exposed their devices to, after the implementation of a countermeasure or a feature. The goal of this research is to develop a characterization of power side channel attacks on different encryption algorithms' implementations to create metrics and methods to evaluate their residual vulnerabilities and added vulnerabilities. This research studies the characteristics that influence the power side leakage, classifies them, and identifies both the residual vulnerabilities and the added vulnerabilities. Residual vulnerabilities are defined as the traits that leave the implementation of the algorithm still vulnerable to power Side Channel Attacks (SCA), sometimes despite the attempt at implementing countermeasures by the designers. Added vulnerabilities to power SCA are defined as vulnerabilities created or enhanced by the algorithm implementations and/or modifications.

The three buckets in which we categorize the encryption algorithm implementations are:

i. Countermeasures against power side channel attacks,

ii. IC power delivery network impact to power leakage (including voltage regulators),

iii. Lightweight ciphers and applications for the Internet of Things (IoT )

From the characterization of masking countermeasures, an example outcome developed is that masking schemes, when uniformly distributed random masks are used, are still vulnerable to collision power attacks. Another example outcome derived is that masked AES, when glitches occur, is still vulnerable to Differential Power Analysis (DPA). We have developed a characterization of power side-channel attacks on the hardware implementations of different symmetric encryption algorithms to provide a detailed analysis of the effectiveness of state-of-the-art countermeasures against local and remote power side-channel attacks. The characterization is accomplished by studying the attributes that influence power side-channel leaks, classifying them, and identifying both residual vulnerabilities and added vulnerabilities. The evaluated countermeasures include masking, hiding, and power delivery network scrambling. But, vulnerability to DPA depends largely on the quality of the leaked power, which is impacted by the characteristics of the device power delivery network.

Countermeasures and deterrents to power side-channel attacks targeting the alteration or scrambling of the power delivery network have been shown to be effective against local attacks where the malicious agent has physical access to the target system. However, remote attacks that capture the leaked information from within the IC power grid are shown herein to be nonetheless effective at uncovering the secret key in the presence of these countermeasures/deterrents. Theoretical studies and experimental analysis are carried out to define and quantify the impact of integrated voltage regulators, voltage noise injection, and integration of on-package decoupling capacitors for both remote and local attacks. An outcome yielded by the studies is that the use of an integrated voltage regulator as a countermeasure is effective for a local attack. However, remote attacks are still effective and hence break the integrated voltage regulator countermeasure. From experimental analysis, it is observed that within the range of designs' practical values, the adoption of on-package decoupling capacitors provides only a 1.3x increase in the minimum number of traces required to discover the secret key. However, the injection of noise in the IC power delivery network yields a 37x increase in the minimum number of traces to discover. Thus, increasing the number of on-package decoupling capacitors or the impedance between the local probing site and the IC power grid should not be relied on as countermeasures to power side-channel attacks, for remote attack schemes. Noise injection should be considered as it is more effective at scrambling the leaked signal to eliminate sensitive identifying information. However, the analysis and experiments carried out herein are applied to regular symmetric ciphers which are not suitable for protecting Internet of Things (IoT) devices.

The protection of communications between IoT devices is of great concern because the information exchanged contains vital sensitive data. Malicious agents seek to exploit those data to extract secret information about the owners or the system. Power side channel attacks are of great concern on these devices because their power consumption unintentionally leaks information correlatable to the device's secret data. Several studies have demonstrated the effectiveness of authenticated encryption with advanced data (AEAD), in protecting communications with these devices. In this research, we have proposed a comprehensive evaluation of the ten algorithm finalists of the National Institute of Standards and Technology (NIST) IoT lightweight cipher competition. The study shows that, nonetheless, some still present some residual vulnerabilities to power side channel attacks (SCA). For five ciphers, we propose an attack methodology as well as the leakage function needed to perform correlation power analysis (CPA). We assert that Ascon, Sparkle, and PHOTON-Beetle security vulnerability can generally be assessed with the security assumptions "Chosen ciphertext attack and leakage in encryption only, with nonce-misuse resilience adversary (CCAmL1)" and "Chosen ciphertext attack and leakage in encryption only with nonce-respecting adversary (CCAL1)", respectively. However, the security vulnerability of GIFT-COFB, Grain, Romulus, and TinyJambu can be evaluated more straightforwardly with publicly available leakage models and solvers. They can also be assessed simply by increasing the number of traces collected to launch the attack.

Rights

© 2023 Aurelien Tchoupou Mozipo

In Copyright. URI: http://rightsstatements.org/vocab/InC/1.0/ This Item is protected by copyright and/or related rights. You are free to use this Item in any way that is permitted by the copyright and related rights legislation that applies to your use. For other uses you need to obtain permission from the rights-holder(s).

Persistent Identifier

https://archives.pdx.edu/ds/psu/40834

Download

Share

COinS