LATIDIA · Ciberseguridad

JASPER: Sesión Especial sobre la Evaluación Conjunta de Confiabilidad y Seguridad de SPlit Computing para la Robustez de Borde

arXiv: 2610.04396v1Tipo de anuncio: nuevo Resumen: Split Computing (SC) permite la implementación eficiente de redes neuronales profundas (DNN) mediante la partición de la inferencia entre los dispositivos de borde y los servidores en la nube. Sin embargo, fea intermedia

La noticia

arXiv:2610.04396v1 Announce Type: new 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