Improving the Detection of Hardware Trojan Horses in Microprocessors via Hamming Codes

Software-exploitable Hardware Trojan Horses (HTHs) can be inserted into commercial microprocessors allowing the attackers to run their own software or to gain unauthorized privileges. As a consequence, HTHs should nowadays be considered a serious threat not only by the academy but also by the indust...

Full description

Saved in:

Bibliographic Details
Published in:	Proceedings of the ... IEEE International Symposium on Defect and Fault Tolerance in VLSI and Nanotechnology Systems (Online) pp. 1 - 6
Main Authors:	Palumbo, Alessandro, Cassano, Luca, Reviriego, Pedro, Ottavi, Marco
Format:	Conference Proceeding
Language:	English
Published:	IEEE 03.10.2023
Subjects:	Benchmark testing Hamming Codes Hardware Security Hardware Trojan Horses Microprocessor-based System Microprocessors Power demand RISC-V Runtime Software Table lookup Very large scale integration
ISSN:	2765-933X
Online Access:	Get full text
Tags:	Add Tag No Tags, Be the first to tag this record!

Description
Summary:	Software-exploitable Hardware Trojan Horses (HTHs) can be inserted into commercial microprocessors allowing the attackers to run their own software or to gain unauthorized privileges. As a consequence, HTHs should nowadays be considered a serious threat not only by the academy but also by the industry. In this paper we present a hardware security checker for the detection of the runtime activation of HTHs. In particular, we aim at detecting HTHs that alter the expected execution flow by launching a malicious program. To achieve this goal the proposed checker is connected between the microprocessor and the main memory and observes the fetching activity. We integrated the proposed checker within a case study based on a RISC-V microprocessor running a set of software benchmarks. The experiment demonstrated that our checker is able to detect 100% of possible HTHs activations with no false alarms. We measured an area overhead of less than 1% w.r.t. LUTs and FFs with 8.5 up to 9.5 BRAM blocks required, a 2.51% power consumption increase, and no working frequency reduction.
ISSN:	2765-933X
DOI:	10.1109/DFT59622.2023.10313563