Content area

Abstract

The work shown in this study demonstrates how Evolutionary Computing (EC) can be used to add trust to Hardware Design Language (HDL) Intellectual Property (IP). HDL IP is often obtained through a 3rd party source due to time and cost constraints, in turn the IP is then considered untrusted by designers. These 3rd party IP could be infected with malicious additions, like Hardware Trojans (HT), or other damaging modifications. HT can often go undetected through standard detection techniques, but even if a designer can identify that there is something wrong with their design, how do they go about repairing it? We propose a study to investigate the ability to remove HT, investigate the use of partial test cases for evolution, and comment on the scalability of the approach. The authors then propose PyGenP, a Genetic Programming (GP) network written in Python, that allows for fast and quick evolution of HDL programs. A Hybrid Memetic GP algorithms that modify the population initialization function is then shown to offer an improvement over traditional GP, while generating better low-order schemas. Finally, we propose an algorithm, using this Hybrid Memetic Genetic Programming initialization function, to perform Targeted Evolution, on select portions of am HDL program, and comment on the improvements the algorithm offers over traditional GP. The authors then close by giving a retrospect of the work completed, and offer recommendations for future work. 

Details

1010268
Business indexing term
Title
Increasing Security and Trust in HDL IP Through Evolutionary Computing
Number of pages
183
Publication year
2022
Degree date
2022
School code
0045
Source
DAI-B 84/8(E), Dissertation Abstracts International
ISBN
9798374401813
Advisor
Committee member
Gallagher, John; Kebede, Temsegen; Kapp, David; Jone, Wen-Ben
University/institution
University of Cincinnati
Department
Engineering and Applied Science: Computer Science and Engineering
University location
United States -- Ohio
Degree
Ph.D.
Source type
Dissertation or Thesis
Language
English
Document type
Dissertation/Thesis
Dissertation/thesis number
30377703
ProQuest document ID
2781100594
Document URL
https://www.proquest.com/dissertations-theses/increasing-security-trust-hdl-ip-through/docview/2781100594/se-2?accountid=208611
Copyright
Database copyright ProQuest LLC; ProQuest does not claim copyright in the individual underlying works.
Database
ProQuest One Academic