TGSM: Towards trustworthy group-based service management for social IoT

Bahareh Farahbakhsh, Ali Fanian, Mohammad Hossein Manshaei

Research output: Contribution to journalArticlepeer-review

15 Scopus citations

Abstract

The Social Internet of Things has been introduced as a social oriented approach to deal with IoT challenges. A service discovery process is also transformed by this paradigm. Since the Social Internet of Things includes millions of services across the network, the question arises that to what extent can we trust service providers? The appropriate selection of trustworthy service provider plays a key role in successful deployment of the SIoT. In this work, we address this issue by proposing a trustworthy group-based service management model (TGSM). The proposed discovery and selection process of an appropriate service provider relies on both trustworthiness measurement and performance assessment. In the trustworthiness measurement case, a punishment mechanism is adopted in trustworthiness evaluation to deal with objects’ selfish behavior. Likewise, we employed the HITS algorithm to propose a global dynamic reputation evaluation approach. We propose a service-based grouping approach as a novel architecture for SIoT network. We also consider the dynamic evolving of social relationships among the SIoT devices. Our SIoT paradigm, due to its intrinsically multilayered nature, is modeled upon a multiplex network. In the performance measurement case, some metrics are used to asses objects’ performance. To evaluate the proposed trust model, we extend the CRAWDAD dataset, and the objects in the dataset are categorized into five types. The simulation results show that the proportions of four selfish types of objects are dominated by completely trustworthy objects, which proves the effectiveness of our proposed TGSM model. Moreover, the results demonstrate HITS centrality measurement's effectiveness, which leads the proportion of trustworthy service providers to nearly double in comparison with degree and eigenvector centrality in this context.

Original languageEnglish (US)
Article number100312
JournalInternet of Things (Netherlands)
Volume13
DOIs
StatePublished - Mar 2021
Externally publishedYes

Keywords

  • HITS Algorithm
  • Multiplex network
  • Punishment mechanism
  • Service based grouping
  • Social internet of things
  • Trust

ASJC Scopus subject areas

  • Artificial Intelligence
  • Computer Science Applications
  • Information Systems
  • Software
  • Hardware and Architecture
  • Computer Science (miscellaneous)
  • Management of Technology and Innovation
  • Engineering (miscellaneous)

Fingerprint

Dive into the research topics of 'TGSM: Towards trustworthy group-based service management for social IoT'. Together they form a unique fingerprint.

Cite this