Logo
ScienceStack
Table of Contents
Fetching ...
Sign In

Security layers and related services within the Horizon Europe NEUROPULS project

Fabio Pavanello, Cedric Marchand, Paul Jimenez, Xavier Letartre, Ricardo Chaves, Niccolò Marastoni, Alberto Lovato, Mariano Ceccato, George Papadimitriou, Vasileios Karakostas, Dimitris Gizopoulos, Roberta Bardini, Tzamn Melendez Carmona, Stefano Di Carlo, Alessandro Savino, Laurence Lerch, Ulrich Ruhrmair, Sergio Vinagrero Gutierrez, Giorgio Di Natale, Elena Ioana Vatajelu

TL;DR

The work presents a photonic-PUF–driven security stack for a neuromorphic photonic accelerator intended for edge computing. It advances weak/strong PUF architectures on a PIC, quality-improvement techniques, and three concrete services—mutual authentication, remote software attestation, and hardware-assisted NN encryption—to enable low-power, robust security. It also analyzes attack resistance and lays out a system-level gem5 simulation framework to quantify performance, reliability, and security under realistic operating conditions. Collectively, the approach enables secure, scalable hardware–software co-design for photonics-enabled edge AI while addressing ML and side-channel threats. The framework aims to deliver practical security primitives with strong binding between PIC and ASIC and verifiable integrity for edge deployments.

Abstract

In the contemporary security landscape, the incorporation of photonics has emerged as a transformative force, unlocking a spectrum of possibilities to enhance the resilience and effectiveness of security primitives. This integration represents more than a mere technological augmentation; it signifies a paradigm shift towards innovative approaches capable of delivering security primitives with key properties for low-power systems. This not only augments the robustness of security frameworks, but also paves the way for novel strategies that adapt to the evolving challenges of the digital age. This paper discusses the security layers and related services that will be developed, modeled, and evaluated within the Horizon Europe NEUROPULS project. These layers will exploit novel implementations for security primitives based on physical unclonable functions (PUFs) using integrated photonics technology. Their objective is to provide a series of services to support the secure operation of a neuromorphic photonic accelerator for edge computing applications.

Security layers and related services within the Horizon Europe NEUROPULS project

TL;DR

The work presents a photonic-PUF–driven security stack for a neuromorphic photonic accelerator intended for edge computing. It advances weak/strong PUF architectures on a PIC, quality-improvement techniques, and three concrete services—mutual authentication, remote software attestation, and hardware-assisted NN encryption—to enable low-power, robust security. It also analyzes attack resistance and lays out a system-level gem5 simulation framework to quantify performance, reliability, and security under realistic operating conditions. Collectively, the approach enables secure, scalable hardware–software co-design for photonics-enabled edge AI while addressing ML and side-channel threats. The framework aims to deliver practical security primitives with strong binding between PIC and ASIC and verifiable integrity for edge deployments.

Abstract

In the contemporary security landscape, the incorporation of photonics has emerged as a transformative force, unlocking a spectrum of possibilities to enhance the resilience and effectiveness of security primitives. This integration represents more than a mere technological augmentation; it signifies a paradigm shift towards innovative approaches capable of delivering security primitives with key properties for low-power systems. This not only augments the robustness of security frameworks, but also paves the way for novel strategies that adapt to the evolving challenges of the digital age. This paper discusses the security layers and related services that will be developed, modeled, and evaluated within the Horizon Europe NEUROPULS project. These layers will exploit novel implementations for security primitives based on physical unclonable functions (PUFs) using integrated photonics technology. Their objective is to provide a series of services to support the secure operation of a neuromorphic photonic accelerator for edge computing applications.
Paper Structure (10 sections, 4 figures, 1 table)

This paper contains 10 sections, 4 figures, 1 table.

Table of Contents

  1. Introduction
  2. Security Primitives and Related Metrics
  3. PUFs
  4. Techniques for PUF quality improvement
  5. Proposed Services and Related Security Protocols
  6. Mutual authentication
  7. Software attestation
  8. Neural network configuration and data encryption
  9. Strengths Against Attacks
  10. System-Level Modeling and Benchmarking

Figures (4)

  • Figure 1: Hardware-Software communication flow for security services in NEUROPULS. Weak and strong PUFs target different security services and supporting circuitry.
  • Figure 2: Schematic of the PUF operation considered in NEUROPULS. OM: optical modulator, PDs: photodiodes. The passive PUF architecture section separates the initial light beam in several different paths and scrambles them before the output. No active devices are present. In the final demonstrator, all the PIC components shown and ASIC-related PUF circuitry will constitute the actual PUF.
  • Figure 3: Relationship between bit aliasing and reliability taking into account the counter threshold
  • Figure 4: Session $i$ of the mutual authentication protocol.