Immune optimization of task scheduling on multidimensional QoS constraints

Abstract

Aiming at the sensitive issues of service quality in cloud computing, a task scheduling tactic with multidimensional QoS constraints is studied. Based on cluster service and user QoS preference, this article constructs an immune optimization model to make a description through formulas and quantify the performance constraints; the utility function of multidimensional QoS is given and then the immune optimization operation is performed with the antibodies. It is beneficial to increase the prediction accuracy of the equality evaluation, and the search for a Pareto optimal set of multiobjective optimization problems is implemented. Finally, the optimum node distribution structure with the highest utility value is obtained. It's shown that the approach gives sufficient consideration of multidimensional user QoS requirements. The results from the test show a significant improvement in average rate of equipment utilization, service time and response time compared to similar algorithms.