QoS adaptation for publish/subscribe middleware in real-time dynamic environments

Automatic Quality of Service (QoS) adaptation is promising approaches in developing real-time systems built on publish/subscribe middleware, and its importance increased when developing huge real-time distributed systems in dynamic environments. In this paper we proposed a new approach by building a closed loop QoS adaptive control for adapting some QoS polices automatically in publish/subscribe middleware. We use some techniques of artificial intelligence to automate QoS adaptation and improve the performance of the real-time systems in dynamic environments. The proposed approach improves the performance of the real-time system by saving its computing capabilities and making it more stable. Experimental results shown in this paper illustrate the effectiveness of the proposed approach.