Securing SoC Processors by Routing Memory Transactions Through FPGA Fabric

This paper explores a radical approach to mitigating unintended software changes, persistent malicious implants, and zero-day exploits on System-on-Chip (SoC) architectures, such as the Xilinx UltraScale MPSoC. The approach creates a hardware root-of-trust by routing all memory transactions through the SoC’s on-chip Field Programmable Gate Array (FPGA) to allow inline monitoring, validation, interception, and corrective techniques. Two distinct formulations are presented and evaluated. The first leverages on-chip Random Access Memory (RAM) directly associated with the SoC’s Processing System (PS); this involves channeling memory transactions through the FPGA and back to the PS using a custom hardware module for address translation. The second leverages off-chip RAM that is exclusively reachable via the FPGA fabric which thereby sits in line between the PS and its memory. Performance results are presented for a wide variety of memory access patterns by running Linux on a ZCU102 development board. This board provides an MPSoC with both RAM configurations, using realistic clock speeds: the multi-core PS operates at 1.2GHz and the associated FPGA fabric operates at 300MHz. The results indicate that the performance overhead typically incurs a 1.5 to 4-fold slowdown that, while significant, represents acceptable values for many mission-critical, high-value applications.

Authors

Institutions

Publication Details

Published
2024-10-28
DOI
https://doi.org/10.1109/milcom61039.2024.10773818
Primary Topic
Physical Unclonable Functions (PUFs) and Hardware Security
Type
article
Field-Weighted Citation Impact
0.00

Funders

Controls
|||
ALL TIME
JAN
FEB
MAR
APR
MAY
JUN
JUL
AUG
SEP
OCT
article

Securing SoC Processors by Routing Memory Transactions Through FPGA Fabric

Stephen Taylor, Jason Dahlstrom, James P. Brock, Ryan Longwell et al.
Physical Unclonable Functions (PUFs) and Hardware Security
article

Securing SoC Processors by Routing Memory Transactions Through FPGA Fabric

Stephen Taylor, Jason Dahlstrom, James P. Brock, Ryan Longwell, Abbie Cliche
article en

Abstract

This paper explores a radical approach to mitigating unintended software changes, persistent malicious implants, and zero-day exploits on System-on-Chip (SoC) architectures, such as the Xilinx UltraScale MPSoC. The approach creates a hardware root-of-trust by routing all memory transactions through the SoC’s on-chip Field Programmable Gate Array (FPGA) to allow inline monitoring, validation, interception, and corrective techniques. Two distinct formulations are presented and evaluated. The first leverages on-chip Random Access Memory (RAM) directly associated with the SoC’s Processing System (PS); this involves channeling memory transactions through the FPGA and back to the PS using a custom hardware module for address translation. The second leverages off-chip RAM that is exclusively reachable via the FPGA fabric which thereby sits in line between the PS and its memory. Performance results are presented for a wide variety of memory access patterns by running Linux on a ZCU102 development board. This board provides an MPSoC with both RAM configurations, using realistic clock speeds: the multi-core PS operates at 1.2GHz and the associated FPGA fabric operates at 300MHz. The results indicate that the performance overhead typically incurs a 1.5 to 4-fold slowdown that, while significant, represents acceptable values for many mission-critical, high-value applications.

Dartmouth College (US)
Advanced Research Projects Agency
Affordable and clean energy
Openalex Percentile: Top 76%
Physical Unclonable Functions (PUFs) and Hardware Security
AI Navigator

Ask Laika to Summarize, Analyze, and Connect papers live on the map.

Summarize Papers & Methodologies

Extract key findings, datasets, and comparative methods across publications.

Benchmark Rankings & Visual Analytics

Rank top research institutions, authors, funders, topics, and journals by Field-Weighted Citation Impact (FWCI) and paper volume with instant charts.

Connect Distant Disciplines

Bridge topological clusters on the map to find hidden collaborative intersections.