JASPER: Special Session on Joint Reliability And Security Assessment of SPlit Computing for Edge Robustness

2026-10-03

AI Summary

This study addresses the robustness blind spots in existing edge partitioned computing evaluations, which typically assess hardware faults and adversarial attacks independently. We propose a joint reliability-security evaluation framework that quantifies intermediate feature vulnerability through neuron-level fault injection and feature map-aware adversarial attack simulation. Furthermore, we introduce a novel joint vulnerability scoring mechanism alongside a confidence-aware extension strategy to uncover deep fragility trends obscured by isolated analyses. Experimental evaluations on ResNet-50 using the ILSVRC-2012 dataset demonstrate that the proposed framework achieves a Mean Relative Accuracy Drop (MRAD) of 44.3%–61.2% and an Attack Success Rate (ASR) of up to 98.8%. These findings effectively guide the design of highly robust edge computing systems.

Abstract

Split Computing (SC) enables efficient deployment of Deep Neural Networks (DNNs) by partitioning inference between edge devices and cloud servers. However, intermediate feature representations are simultaneously exposed to hardware faults and adversarial attacks, which are traditionally evaluated independently. This paper presents a unified framework for the joint assessment of reliability and security in Split Computing. First, reliability is characterized through neuron-level fault injection using the Mean Relative Accuracy Degradation (MRAD) while security through feature-map-aware adversarial attacks simulations using the Attack Success Rate (ASR). Based on these complementary analyses, the Joint Vulnerability Score (JVS) is introduced, along with a confidence-aware extension that jointly captures prediction errors and confidence degradation. The framework is evaluated on ten Split Computing configurations based on ResNet-50 trained on ILSVRC-2012. Experimental results show substantial differences across compression strategies, with MRAD ranging from 44.3% to 61.2% under fault injection, while adversarial attacks achieve up to 98.8% ASR. Furthermore, the proposed joint metrics reveal vulnerability trends that remain hidden when reliability and security are analyzed independently, providing a more comprehensive methodology for designing dependable Split Computing systems.

Problem

• Split Computing
• Reliability
• Security
• Adversarial Attacks
• Hardware Faults

Innovation

• Split Computing
• Joint Reliability and Security Assessment
• Joint Vulnerability Score (JVS)
• Adversarial Attacks
• Fault Injection