基于不精确计算移动实时数据库服务质量管理

移动实时数据库服务应用逐渐广泛，但系统负载不可预测和有限资源常导致事务重启或夭折，给系统带来损失甚至灾难。传统的基于最长执行时间实时调度算法已不能满足性能要求，提出结合不精确计算和反馈控制的新算法。考虑更新数据对象间联系并结合时间和值域有效性，提出性能新参数标准去保证系统性能和服务质量。通过仿真实验表明：算法可从稳定性能和暂态性能保证系统预定的服务质量规范。

QoS management for real-time and mobile database systems based on imprecise computation

The demand on mobile and real-time database application and service becomes increasingly extensive, so that competing for the limited resources and predicting workload imprecisely lead to abortion or restart for system even catastrophe. Because a real-time process scheduler using worst-case execution time is inefficient for optimizing performance, a scheduling algorithm for update transactions based on imprecise computation and feedback control loop was presented in this paper. A performance metrics based on validity in time domain and value domain was proposed to ensure the quality of service in mobile and real-time database systems. The algorithms and feedback control architecture can guarantee the transient and steady performance of Quality of Service (QoS), which is verified by simulation experiments.