DyST: Dynamic Specification Mining for Heterogenous IoT Systems with WoT

Ege Korkan, Silvia Oliva Ramirez, Sebastian Steinhorst

Research output: Chapter in Book/Report/Conference proceedingConference contributionpeer-review

Abstract

The comprehension of a distributed system and its verification is one of the most challenging problems in today’s software engineering, commonly referred to as observability. The complexity increases when one cannot control all the components, like in IoT systems composed of third-party devices. The Web of Things standards by the W3C help with this by describing what one can do with an IoT device via network messages. However, no work has leveraged these standards to offer an observability solution that works with any set of IoT devices. This work addresses this gap by proposing a method to verify the correctness of the system by mining its specification from device interactions. Our approach can reverse engineer complex application logic in the form of UML Sequence Diagrams from the analysis of network messages of any protocol between the devices during system runtime, which can be used to programmatically assert the correctness of the mined specification. We have evaluated our approach with three case studies to assess our mining technique, the performance of our algorithms, and the applicability of our contributions to system verification in the IoT. Our results show that our approach can produce accurate Sequence Diagrams that help understand and verify the behavior of IoT systems.

Original languageEnglish
Title of host publicationWeb Engineering - 24th International Conference, ICWE 2024, Proceedings
EditorsKostas Stefanidis, Kari Systä, Maristella Matera, Sebastian Heil, Haridimos Kondylakis, Elisa Quintarelli
PublisherSpringer Science and Business Media Deutschland GmbH
Pages147-162
Number of pages16
ISBN (Print)9783031623615
DOIs
StatePublished - 2024
Event24th International Conference on Web Engineering, ICWE 2024 - Tampere, Finland
Duration: 17 Jun 202420 Jun 2024

Publication series

NameLecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics)
Volume14629 LNCS
ISSN (Print)0302-9743
ISSN (Electronic)1611-3349

Conference

Conference24th International Conference on Web Engineering, ICWE 2024
Country/TerritoryFinland
CityTampere
Period17/06/2420/06/24

Keywords

  • Internet of Things
  • Process Mining
  • Web of Things

Fingerprint

Dive into the research topics of 'DyST: Dynamic Specification Mining for Heterogenous IoT Systems with WoT'. Together they form a unique fingerprint.

Cite this