Title: A GA-based approach with an interval VIKOR method for solving the constrained QoS-aware service composition in dynamic IoT environments

Authors: Fateh Seghir; Renda Kouachi

Addresses: Intelligent Systems Laboratory (LSI), Faculty of Technology, Sétif 1 University, Sétif, Algeria ' Department of Computer Science, Faculty of Sciences, Sétif 1 University, Sétif, Algeria

Abstract: The QoS-aware service composition (QSC) with global QoS user-constraints in dynamic IoT environments is an NP-hard problem, where the QoS values of the IoT services are often ambiguous in nature due to various reasons such as network topology changes, IoT devices mobility and economic policies. Therefore, motivated by the fact that the interval number is an efficient and a simple model to express the imprecision of the QoS properties; the QSC in uncertain IoT environments is formulated as an interval multi-criteria optimisation (INQSC) problem. Furthermore, to solve the modelled INQSC, we provide a GA-based optimisation approach, which integrates an interval VIKOR method to deal with feasible solutions ranking, an interval QoS constraint violation sorting to rank infeasible solutions, and a local search operator with an elitism replacement to enhance both the exploitation and the exploration abilities of the provided optimisation approach. The experimental comparison of our proposal with a recently provided GAP approach demonstrates the performance and the effectiveness of the proposed GA-based approach.

Keywords: IoT services; quality of service; QoS; uncertainty; interval number; multi-criteria optimisation; VIKOR method; genetic algorithm.

DOI: 10.1504/IJWGS.2022.123656

International Journal of Web and Grid Services, 2022 Vol.18 No.3, pp.297 - 336

Received: 14 Jun 2020
Accepted: 12 May 2021

Published online: 30 Jun 2022 *

Full-text access for editors Full-text access for subscribers Purchase this article Comment on this article