PORTALE DELLA DIDATTICA

Ricerca CERCA
  KEYWORD

Malicious Code Detection for Automotive-secured systems

keywords AUTHENTICATION, AUTOMOTIVE, INTRUSION DETETCTION, MALWARE, SECURE BOOT, SECURITY

Reference persons STEFANO DI CARLO, ALESSANDRO SAVINO

External reference persons OBERT FRANCO

Research Groups DAUIN - GR-24 - SMILIES - reSilient coMputer archItectures and LIfE Sci

Thesis type APPLIED, APPLIED RESEARCH, EXPERIMENTAL RESEARCH, HARDWARE DESIGN

Description The recent period is growing the number of attacks that leverage on relaying third parties’ software exploits. Many security vulnerability detection tools are already available today in the market. These tools allow scanning of the source code of projects, infrastructure and applications and reporting potential security vulnerabilities and weaknesses. The current approach swing between two categories of analyzers: dynamic and static. Both are not useful when the third-party software module is delivered to the customer directly compiled code without source code sharing. This habit is quite common in the automotive domain for protecting Intellectual Properties (IP), made easy by AUTOSAR Architecture Framework that guarantees software module compatibility.
In the present thesis, the student shall develop a novel static analysis based on traditional and deep-learning methods that detect potential software vulnerabilities on binary code instead of traditional source code.

Notes In collaboration with PUNCH Softronix


Deadline 28/11/2023      PROPONI LA TUA CANDIDATURA