Clustering-based and QoS-aware services composition algorithm for ambient intelligence

Due to the dynamic nature of ubiquitous computing and ambient intelligence (AmI) environments, a challenging issue that needs to be addressed is how to construct composite services that satisfy users’ requirements in terms of quality of service (QoS). In this paper, a clustering-based and QoS-aware...

Ausführliche Beschreibung

Gespeichert in:
Bibliographische Detailangaben
Veröffentlicht in:Information sciences Jg. 482; S. 419 - 439
Hauptverfasser: Khanouche, Mohamed Essaid, Attal, Ferhat, Amirat, Yacine, Chibani, Abdelghani, Kerkar, Moussa
Format: Journal Article
Sprache:Englisch
Veröffentlicht: Elsevier Inc 01.05.2019
Elsevier
Schlagworte:
Ambient intelligence
Artificial Intelligence
Computer Science
k-means clustering method
QoS constraints
Quality of service
Services composition
QoS constraints
Services composition
Ambient intelligence
Quality of service
k-means clustering method
ISSN:0020-0255, 1872-6291
Online-Zugang:Volltext
Tags: Tag hinzufügen
Keine Tags, Fügen Sie den ersten Tag hinzu!
Beschreibung
Zusammenfassung:Due to the dynamic nature of ubiquitous computing and ambient intelligence (AmI) environments, a challenging issue that needs to be addressed is how to construct composite services that satisfy users’ requirements in terms of quality of service (QoS). In this paper, a clustering-based and QoS-aware services composition algorithm (CQCA) is proposed. To increase the composition optimality and reduce the composition time, the candidate services are first partitioned into clusters, where each cluster represents a QoS level. In addition, a new formulation of the utility function based on the use of the characteristics of the resulting clusters is proposed to remove unpromising candidate services in terms of QoS. A lexicographic optimization method is then exploited to filter out candidate services that have low QoS attributes values. Finally, a search tree is constructed to find near-to-optimal compositions. The obtained performance shows that the proposed algorithm outperforms other composition approaches by finding very near-to-optimal compositions in a reduced composition time.
ISSN:0020-0255
1872-6291
DOI:10.1016/j.ins.2019.01.015