A software-defined architecture for control of IoT cyberphysical systems

Ala’ Darabseh, Nikolaos M. Freris

Research output: Contribution to journalArticle

Abstract

Based on software-defined principles, we propose a holistic architecture for cyberphysical systems (CPS) and internet of things (IoT) applications, and highlight the merits pertaining to scalability, flexibility, robustness, interoperability, and cyber security. Our design especially capitalizes on the computational units possessed by smart agents, which may be utilized for decentralized control and in-network data processing. We characterize the data flow, communication flow, and control flow that assimilate a set of components such as sensors, actuators, controllers, and coordinators in a systemic programmable fashion. We specifically aim for distributed and decentralized decision-making by spreading the control over several hierarchical layers. In addition, we propose a middleware layer to encapsulate units and services for time-critical operations in highly dynamic environments. We further enlist a multitude of vulnerabilities to cyberattacks, and integrate software-defined solutions for enabling resilience, detection and recovery. In this purview, several controllers cooperate to identify and respond to security threats and abnormal situations in a self-adjusting manner. Last, we illustrate numerical simulations in support of the virtues of a software-defined design for CPS and IoT.

Original languageEnglish (US)
Pages (from-to)1107-1122
Number of pages16
JournalCluster Computing
Volume22
Issue number4
DOIs
StatePublished - Dec 1 2019

Keywords

  • Cyber security
  • Cyberphysical systems (CPS)
  • Decentralized control
  • Distributed systems
  • Internet of things (IoT)
  • Middleware
  • Software defined systems (SDSys)

ASJC Scopus subject areas

  • Software
  • Computer Networks and Communications

Fingerprint Dive into the research topics of 'A software-defined architecture for control of IoT cyberphysical systems'. Together they form a unique fingerprint.

Cite this