Abstract
Side channel attacks, such as Differential Power Analysis (DPA), denote a special class of attacks in which sensitive key information is unveiled through information extracted from the physical device executing a cryptographic algorithm. This information leakage, known as side channel information, occurs from computations in a non-ideal system composed of electronic devices such as transistors. Power dissipation is one classic side channel source, which relays information of the data being processed. DPA uses statistical analysis to identify data-dependent correlations in sets of power measurements.
Countermeasures against DPA focus on hiding or masking techniques at different levels of design abstraction and are typically associated with high power and area cost. Emerging technologies such as Resistive Random Access Memory (RRAM), offer unique opportunities to mitigate DPAs with their inherent memristor device characteristics such as variability in write time, ultra low power (0.1-3 pJ/bit), and high density (4F2).
In this research, an RRAM based architecture is proposed to mitigate the DPA attacks by obfuscating the power profile. Specifically, a dual RRAM based memory module masks the power dissipation of the actual transaction by accessing both the data and its complement from the memory in tandem. DPA attack resiliency for a 128-bit AES cryptoprocessor using RRAM and CMOS memory modules is compared against baseline CMOS only technology.
In the proposed AES architecture, four single port RRAM memory units store the
intermediate state of the encryption. The correlation between the state data and sets of power measurement is masked due to power dissipated from inverse data access on dual RRAM memory. A customized simulation framework is developed to design the attack scenarios using Synopsys and Cadence tool suites, along with a Hamming weight DPA attack module. The attack mounted on a baseline CMOS architecture is successful and the full key is recovered. However, DPA attacks mounted on the dual CMOS and RRAM based AES cryptoprocessor yielded unsuccessful results with no keys recovered, demonstrating the resiliency of the proposed architecture against DPA attacks.
Library of Congress Subject Headings
Nonvolatile random-access memory; Memristors; Computer crimes--Prevention; Data encryption (Computer science); Low voltage integrated circuits
Publication Date
8-2013
Document Type
Thesis
Student Type
Graduate
Degree Name
Computer Engineering (MS)
Department, Program, or Center
Computer Engineering (KGCOE)
Advisor
Dhireesha Kudithipudi
Advisor/Committee Member
Marcin Łukowiak
Advisor/Committee Member
Amlan Ganguly
Recommended Citation
Khedkar, Ganesh Chandrakantrao, "Power Profile Obfuscation using RRAMs to Counter DPA Attacks" (2013). Thesis. Rochester Institute of Technology. Accessed from
https://repository.rit.edu/theses/8373
Campus
RIT – Main Campus
Plan Codes
CMPE-MS
Comments
Physical copy available from RIT's Wallace Library at TK7895.M4 K44 2013