Towards the Integration of Reliability and Security Mechanisms to Enhance the Fault Resilience of Neural Networks

Nowadays, many electronic systems store valuable Intellectual Property (IP) information inside Non-Volatile Memories (NVMs). Designers widely use encryption mechanisms to enhance the integrity of such IPs and protect them from any unauthorized access or modification. At the same time, often such IPs...

Descripción completa

Guardado en:

Detalles Bibliográficos
Autores principales:	Nikolaos Ioannis Deligiannis, Riccardo Cantoro, Matteo Sonza Reorda, Marcello Traiola, Emanuele Valea
Formato:	article
Lenguaje:	EN
Publicado:	IEEE 2021
Materias:	Artificial neural networks convolutional neural networks encryption fault detection fault injection non-volatile memories Electrical engineering. Electronics. Nuclear engineering TK1-9971
Acceso en línea:	https://doaj.org/article/ce17281b886a4d419a89839ecce4c9ba
Etiquetas:	Agregar Etiqueta Sin Etiquetas, Sea el primero en etiquetar este registro!

id	oai:doaj.org-article:ce17281b886a4d419a89839ecce4c9ba
record_format	dspace
spelling	oai:doaj.org-article:ce17281b886a4d419a89839ecce4c9ba2021-12-01T00:00:48ZTowards the Integration of Reliability and Security Mechanisms to Enhance the Fault Resilience of Neural Networks2169-353610.1109/ACCESS.2021.3129149https://doaj.org/article/ce17281b886a4d419a89839ecce4c9ba2021-01-01T00:00:00Zhttps://ieeexplore.ieee.org/document/9618910/https://doaj.org/toc/2169-3536Nowadays, many electronic systems store valuable Intellectual Property (IP) information inside Non-Volatile Memories (NVMs). Designers widely use encryption mechanisms to enhance the integrity of such IPs and protect them from any unauthorized access or modification. At the same time, often such IPs are critical from a reliability standpoint. Thus, dedicated techniques are employed to detect possible reliability threats (e.g., transient faults affecting the NVM content). The weights of a neural network (NN) model (e.g., integrated into an object detection system for autonomous driving or robotics) are typical examples of precious IP items. Indeed, NN weights often constitute proprietary data, stemming from an extensive and costly training process; moreover, their correctness is key for the NN to work reliably. In this article, we explore the capability of encryption mechanisms to ensure protection from both reliability threats. In particular, we assess, via extensive fault injection campaigns, the capability of different memory encryption schemes – usually used only for security purposes – to detect faults and thus, enhance the reliability of the system. Experimental results show that, by cleverly choosing the proper encryption scheme, it is possible to achieve very high fault detection rates (greater than 99%) with respect to Multiple Bit Upsets. The gathered results pave the way to the integration of reliability and security mechanisms to achieve better results with lower costs.Nikolaos Ioannis DeligiannisRiccardo CantoroMatteo Sonza ReordaMarcello TraiolaEmanuele ValeaIEEEarticleArtificial neural networksconvolutional neural networksencryptionfault detectionfault injectionnon-volatile memoriesElectrical engineering. Electronics. Nuclear engineeringTK1-9971ENIEEE Access, Vol 9, Pp 155998-156012 (2021)
institution	DOAJ
collection	DOAJ
language	EN
topic	Artificial neural networks convolutional neural networks encryption fault detection fault injection non-volatile memories Electrical engineering. Electronics. Nuclear engineering TK1-9971
spellingShingle	Artificial neural networks convolutional neural networks encryption fault detection fault injection non-volatile memories Electrical engineering. Electronics. Nuclear engineering TK1-9971 Nikolaos Ioannis Deligiannis Riccardo Cantoro Matteo Sonza Reorda Marcello Traiola Emanuele Valea Towards the Integration of Reliability and Security Mechanisms to Enhance the Fault Resilience of Neural Networks
description	Nowadays, many electronic systems store valuable Intellectual Property (IP) information inside Non-Volatile Memories (NVMs). Designers widely use encryption mechanisms to enhance the integrity of such IPs and protect them from any unauthorized access or modification. At the same time, often such IPs are critical from a reliability standpoint. Thus, dedicated techniques are employed to detect possible reliability threats (e.g., transient faults affecting the NVM content). The weights of a neural network (NN) model (e.g., integrated into an object detection system for autonomous driving or robotics) are typical examples of precious IP items. Indeed, NN weights often constitute proprietary data, stemming from an extensive and costly training process; moreover, their correctness is key for the NN to work reliably. In this article, we explore the capability of encryption mechanisms to ensure protection from both reliability threats. In particular, we assess, via extensive fault injection campaigns, the capability of different memory encryption schemes – usually used only for security purposes – to detect faults and thus, enhance the reliability of the system. Experimental results show that, by cleverly choosing the proper encryption scheme, it is possible to achieve very high fault detection rates (greater than 99%) with respect to Multiple Bit Upsets. The gathered results pave the way to the integration of reliability and security mechanisms to achieve better results with lower costs.
format	article
author	Nikolaos Ioannis Deligiannis Riccardo Cantoro Matteo Sonza Reorda Marcello Traiola Emanuele Valea
author_facet	Nikolaos Ioannis Deligiannis Riccardo Cantoro Matteo Sonza Reorda Marcello Traiola Emanuele Valea
author_sort	Nikolaos Ioannis Deligiannis
title	Towards the Integration of Reliability and Security Mechanisms to Enhance the Fault Resilience of Neural Networks
title_short	Towards the Integration of Reliability and Security Mechanisms to Enhance the Fault Resilience of Neural Networks
title_full	Towards the Integration of Reliability and Security Mechanisms to Enhance the Fault Resilience of Neural Networks
title_fullStr	Towards the Integration of Reliability and Security Mechanisms to Enhance the Fault Resilience of Neural Networks
title_full_unstemmed	Towards the Integration of Reliability and Security Mechanisms to Enhance the Fault Resilience of Neural Networks
title_sort	towards the integration of reliability and security mechanisms to enhance the fault resilience of neural networks
publisher	IEEE
publishDate	2021
url	https://doaj.org/article/ce17281b886a4d419a89839ecce4c9ba
work_keys_str_mv	AT nikolaosioannisdeligiannis towardstheintegrationofreliabilityandsecuritymechanismstoenhancethefaultresilienceofneuralnetworks AT riccardocantoro towardstheintegrationofreliabilityandsecuritymechanismstoenhancethefaultresilienceofneuralnetworks AT matteosonzareorda towardstheintegrationofreliabilityandsecuritymechanismstoenhancethefaultresilienceofneuralnetworks AT marcellotraiola towardstheintegrationofreliabilityandsecuritymechanismstoenhancethefaultresilienceofneuralnetworks AT emanuelevalea towardstheintegrationofreliabilityandsecuritymechanismstoenhancethefaultresilienceofneuralnetworks
_version_	1718406165331705856

Towards the Integration of Reliability and Security Mechanisms to Enhance the Fault Resilience of Neural Networks

Ejemplares similares