大规模计算网络并行任务调度模型及其参数方法研究

With wide applications of Internet technology and the rapid development of mobile computing, grid computing, and cloud computing, large-scale computing network system, heterogeneous resources as well as the diversity of user requirements makes the scheduling problem faced by more new challenges. This project researches the model and parametric methods of parallel task scheduling in large-scale computing network system. The main contents include: (1) large-scale computing network system resource management and scheduling mechanism such as the various types of cloud computing and grid computing, according to the various parameters effect the services provider and the service requestor in the scheduling process, construct parallel task scheduling model by abstract method; (2) combined with the classic parallel task scheduling model theory and method, applied parameter computing theories and techniques to explore the complexity of the scheduling problem, develop its feasible approximation algorithm and parameter algorithm; (3) establishment of a scheduling algorithm verification platform for cloud computing simulation environment, a new scheduling algorithm for cloud computing environments provided by experimental improvement. The project aims to improve system performance and utility of the large-scale computing networks, provide the basis for scheduling algorithms for the mainstream networks in parallel and distributed processing.

Internet技术的广泛应用和移动计算、网格计算、特别是云计算的快速发展，网络服务系统的大规模化、资源的异构动态性以及用户需求的多样化使得其调度问题面临更多新的挑战。本项目研究大规模计算网络系统的并行任务调度模型和参数方法。主要内容包括：（1）研究各类云计算、网格计算等大规模计算网络系统的资源管理与调度机制，根据调度所涉及的服务提供方与服务请求方相关的各种影响参数，抽象出相应的并行任务调度模型；（2）结合经典的并行任务调度模型理论与方法，应用参数计算理论与技术，探讨本项目调度问题的复杂性，及其可行近似算法与参数算法。（3）针对云计算仿真环境建立调度算法验证平台，并通过实验改进得到适用于云计算环境的新型调度算法。本项目的研究旨在提高大规模计算网络系统的性能和效用，同时为主流网络并行与分布式处理提供调度算法基础。

Internet技术的广泛应用和移动计算、网格计算、特别是云计算的快速发展，网络服务系统的大规模化、资源的异构动态性以及用户需求的多样化使得其调度问题面临更多新的挑战。本项目研究云计算网络系统，主要内容包括：（1）研究了云计算系统的资源管理与调度机制，建立云计算实验平台；（2）探讨了云计算调度问题近似参数算法与参数近似算法的可行性；（3）研究了云计算环境节能模型与优化；（4）研究了云环境下高维差分进化算法及其优化；（5）继续研究了近似算法的优化。本项目的研究旨在提高大规模计算网络系统的性能和效用，同时为主流网络并行与分布式处理提供调度算法基础。