Byzantine-Tolerant Inference in Distributed Deep Intelligent System: Challenges and Opportunities

Jinhang Choi, Zeinab Hakimi, John Morgan Sampson, Vijaykrishnan Narayanan

Research output: Contribution to journalArticle

Abstract

Cyber-physical systems in the Internet-of-Things (IoT) era increasingly need responses that are not only timely, but also intelligent. To that end, decentralized Deep Neural Network (DNN) systems have been studied for near-sensor processing to enable localized inference and global network partition in a given, limited power budget. In the real world, however, such systems are fundamentally heterogeneous in per-node information quality, since each virtually or physically dispersed sensor captures only part of the observable environment. Recent work in collective DNN systems has leveraged this variation in information quality across nodes in a network hierarchy to improve accuracy and reduce design cost. Unfortunately, in noisy or dynamic environments, there are clear optimization challenges in properly measuring information quality or its proxies, which can make such systems even more sensitive to noise than a quality-agnostic design. This implies that the unequal weighting in an information-quality exploiting distributed DNN will likely present new opportunities to subvert collective decision making. Inspired by the classical fault models of distributed systems, specifically those designed to endure coordinated node failure (Byzantine failure), this article explores interactions between collaborative DNN models, their reliance on information quality metrics in various fault scenarios, and the corresponding system design costs for achieving reasonable accuracy and reliability. As a proof-of-concept, we perform a case study on a distributed multi-view camera system operating under faults introduced both by environmental noise and adversarial inputs, and present results on inference robustness supported by consensus mechanisms in Byzantine settings.

Original languageEnglish (US)
Article number8790811
Pages (from-to)509-519
Number of pages11
JournalIEEE Journal on Emerging and Selected Topics in Circuits and Systems
Volume9
Issue number3
DOIs
StatePublished - Sep 1 2019

Fingerprint

Intelligent systems
Sensors
Costs
Decision making
Cameras
Systems analysis
Deep neural networks
Processing

All Science Journal Classification (ASJC) codes

  • Electrical and Electronic Engineering

Cite this

@article{ee15e7f5851045f1b64286107e930f7c,
title = "Byzantine-Tolerant Inference in Distributed Deep Intelligent System: Challenges and Opportunities",
abstract = "Cyber-physical systems in the Internet-of-Things (IoT) era increasingly need responses that are not only timely, but also intelligent. To that end, decentralized Deep Neural Network (DNN) systems have been studied for near-sensor processing to enable localized inference and global network partition in a given, limited power budget. In the real world, however, such systems are fundamentally heterogeneous in per-node information quality, since each virtually or physically dispersed sensor captures only part of the observable environment. Recent work in collective DNN systems has leveraged this variation in information quality across nodes in a network hierarchy to improve accuracy and reduce design cost. Unfortunately, in noisy or dynamic environments, there are clear optimization challenges in properly measuring information quality or its proxies, which can make such systems even more sensitive to noise than a quality-agnostic design. This implies that the unequal weighting in an information-quality exploiting distributed DNN will likely present new opportunities to subvert collective decision making. Inspired by the classical fault models of distributed systems, specifically those designed to endure coordinated node failure (Byzantine failure), this article explores interactions between collaborative DNN models, their reliance on information quality metrics in various fault scenarios, and the corresponding system design costs for achieving reasonable accuracy and reliability. As a proof-of-concept, we perform a case study on a distributed multi-view camera system operating under faults introduced both by environmental noise and adversarial inputs, and present results on inference robustness supported by consensus mechanisms in Byzantine settings.",
author = "Jinhang Choi and Zeinab Hakimi and Sampson, {John Morgan} and Vijaykrishnan Narayanan",
year = "2019",
month = "9",
day = "1",
doi = "10.1109/JETCAS.2019.2933807",
language = "English (US)",
volume = "9",
pages = "509--519",
journal = "IEEE Journal on Emerging and Selected Topics in Circuits and Systems",
issn = "2156-3357",
publisher = "IEEE Circuits and Systems Society",
number = "3",

}

TY - JOUR

T1 - Byzantine-Tolerant Inference in Distributed Deep Intelligent System

T2 - Challenges and Opportunities

AU - Choi, Jinhang

AU - Hakimi, Zeinab

AU - Sampson, John Morgan

AU - Narayanan, Vijaykrishnan

PY - 2019/9/1

Y1 - 2019/9/1

N2 - Cyber-physical systems in the Internet-of-Things (IoT) era increasingly need responses that are not only timely, but also intelligent. To that end, decentralized Deep Neural Network (DNN) systems have been studied for near-sensor processing to enable localized inference and global network partition in a given, limited power budget. In the real world, however, such systems are fundamentally heterogeneous in per-node information quality, since each virtually or physically dispersed sensor captures only part of the observable environment. Recent work in collective DNN systems has leveraged this variation in information quality across nodes in a network hierarchy to improve accuracy and reduce design cost. Unfortunately, in noisy or dynamic environments, there are clear optimization challenges in properly measuring information quality or its proxies, which can make such systems even more sensitive to noise than a quality-agnostic design. This implies that the unequal weighting in an information-quality exploiting distributed DNN will likely present new opportunities to subvert collective decision making. Inspired by the classical fault models of distributed systems, specifically those designed to endure coordinated node failure (Byzantine failure), this article explores interactions between collaborative DNN models, their reliance on information quality metrics in various fault scenarios, and the corresponding system design costs for achieving reasonable accuracy and reliability. As a proof-of-concept, we perform a case study on a distributed multi-view camera system operating under faults introduced both by environmental noise and adversarial inputs, and present results on inference robustness supported by consensus mechanisms in Byzantine settings.

AB - Cyber-physical systems in the Internet-of-Things (IoT) era increasingly need responses that are not only timely, but also intelligent. To that end, decentralized Deep Neural Network (DNN) systems have been studied for near-sensor processing to enable localized inference and global network partition in a given, limited power budget. In the real world, however, such systems are fundamentally heterogeneous in per-node information quality, since each virtually or physically dispersed sensor captures only part of the observable environment. Recent work in collective DNN systems has leveraged this variation in information quality across nodes in a network hierarchy to improve accuracy and reduce design cost. Unfortunately, in noisy or dynamic environments, there are clear optimization challenges in properly measuring information quality or its proxies, which can make such systems even more sensitive to noise than a quality-agnostic design. This implies that the unequal weighting in an information-quality exploiting distributed DNN will likely present new opportunities to subvert collective decision making. Inspired by the classical fault models of distributed systems, specifically those designed to endure coordinated node failure (Byzantine failure), this article explores interactions between collaborative DNN models, their reliance on information quality metrics in various fault scenarios, and the corresponding system design costs for achieving reasonable accuracy and reliability. As a proof-of-concept, we perform a case study on a distributed multi-view camera system operating under faults introduced both by environmental noise and adversarial inputs, and present results on inference robustness supported by consensus mechanisms in Byzantine settings.

UR - http://www.scopus.com/inward/record.url?scp=85070711905&partnerID=8YFLogxK

UR - http://www.scopus.com/inward/citedby.url?scp=85070711905&partnerID=8YFLogxK

U2 - 10.1109/JETCAS.2019.2933807

DO - 10.1109/JETCAS.2019.2933807

M3 - Article

AN - SCOPUS:85070711905

VL - 9

SP - 509

EP - 519

JO - IEEE Journal on Emerging and Selected Topics in Circuits and Systems

JF - IEEE Journal on Emerging and Selected Topics in Circuits and Systems

SN - 2156-3357

IS - 3

M1 - 8790811

ER -