面向无线城域网的Cloudlet资源调度方法研究

Cloudlet can provide low latency and high bandwidth computation offloading services for mobile terminals. Deploying Cloudlet in Wireless Metropolitan Area Networks (WMAN) has broad application prospects. However, due to limited and heterogeneous resources in Cloudlet, it is challenging to effectively and efficiently respond to the diverse and dynamic requests from the huge user population in a WMAN. Therefore, research on Cloudlet resource scheduling in WMAN is timely and important. In view of this, this proposed project will focus on the cloudlet resource scheduling method in WMAN. Firstly, according to the user location information and mobile characteristics, a Cloudlet placement method that satisfies the minimum user access delay requirement will be proposed to meet the user mobility. Secondly, a multi-user Cloudlet service selection method will be proposed to meet the diverse needs of the user requests. Finally, according to the characteristics of partitioned applications, a multi-objective computational offloading method for multiple users will be developed under the condition of meeting the time constraints.

Cloudlet可以为移动终端提供低时延和高带宽的计算迁移服务，将Cloudlet部署在无线城域网具有很广阔的应用前景。但是，由于无线城域网中具有庞大的用户规模，而Cloudlet资源受限且具有异构性，同时，其需要响应的用户请求具有异构和动态变化特征，所以，无线城域网中Cloudlet资源调度问题面临着严峻挑战。有鉴于此，本项目将深入研究无线城域网中Cloudlet资源调度方法。首先，根据用户位置信息和移动特征，提出一种支持用户移动性，满足用户访问时延最小的Cloudlet放置方法；其次，在满足用户请求多样性需求下，提出一种多用户Cloudlet服务选择方法；最后，根据可划分应用特征，提出一种满足时间约束的多用户多目标计算迁移方法。

Cloudlet可以为移动终端提供低时延和高带宽的计算迁移服务，将Cloudlet部署在无线城域网具有很广阔的应用前景。但是，无线城域网中用户规模庞大，并且Cloudlet资源受限且具有异构性，同时，其需要响应的用户请求具有异构和动态变化特征，所以，无线城域网中Cloudlet资源调度问题面临着严峻挑战。有鉴于此，本项目将深入研究无线城域网中Cloudlet资源调度方法。首先，根据用户位置信息和移动特征，提出一种支持用户移动性，满足用户访问时延最小的Cloudlet放置方法；其次，在满足用户请求多样性需求下，提出一种多用户Cloudlet服务选择方法；最后，根据可划分应用特征，提出一种满足时间约束的多用户多目标计算迁移方法。