Adaptive Web QoS controller based on online system identification using quantum-behaved particle swarm optimization

Wei Fang, Jun Sun, Xiaojun Wu, Vasile Palade

    Research output: Contribution to journalArticle

    12 Citations (Scopus)


    Web quality of service (QoS) control can effectively prevent a Web Server from becoming overloaded. This paper presents an adaptive controller for Web QoS, which dynamically adjusts parameters of the proportional-integral (PI) controller according to the changes of the Web server model. Different from the existing methods based on the off-line system identification, the adaptive Web QoS controller is implemented based on the online system identification through a quantum-behaved particle swarm optimization (QPSO) algorithm and a mutated version of QPSO (MuQPSO). The proposed approach for online system identification by QPSO and MuQPSO shows better performance than genetic algorithms and the particle swarm optimization algorithm on the simulation tests. Then, the performance of the adaptive controller for Web QoS is tested in three experiments and compared with that of fixed PI controller. Experimental results show that the adaptive control employing online system identification by QPSO and MuQPSO algorithms is able to control the Web server resource more effectively in case of overload and, thus, improves the Web QoS.
    Original languageEnglish
    Pages (from-to)1715-1725
    JournalSoft Computing
    Issue number6
    Publication statusPublished - 2014


    Bibliographical note

    This article is not yet available in the repository.


    • Adaptive control
    • Particle swarm optimization
    • Quantum-behaved particle swarm optimization
    • Web QoS

    Cite this