원문정보
초록
영어
Most existing task scheduling algorithmsfail to aware users' QoS preference and result in low user satisfaction rate for they do not reflect users’QoS requirements. We classify QoS factors into four main QoS class which users understand well and can describe their important level, andintroduce AHP method to help user decide the class weight and avoid judgment logical error.Then, we improve existing standard PSO scheduling by use above AHP based different weights for different QoS classes to make PSO haveQoS preference awareness ability.The simulations show our method gets obviously higher user satisfaction rate and maintains the efficiency at the same time.Finally, the simulations also point out that the hierarchical scheduling is necessary to avoid the common tasks take over the special resource.
목차
1. Introduction
2. Related Works
3. The PSO Basedtask Scheduling Introduction
3.1. Standard PSO introduction
3.2. The existing matrix of PSO
3.3. Fitness function definition for multi-QoS
4. AHP Integrated to aware QoS preference
4.1. The analytic hierarchy processintroduction
4.2. The AHP based weight defined for different class tasks
4.3. The weight definition of inner class
4.4. AHP-PSO task scheduling algorithm
5. Simulations and Analysis
5.1. Simulation environment
5.2. Simulation result and analysis
6. Conclusion
Acknowledgments
References