作　　者： (张春光）; (曾广平）; (王洪泊）; (涂序彦）;

机构地区： 北京科技大学计算机与通信工程学院,北京100083

出　　处： 《中北大学学报（自然科学版）》 2017年第3期333-340,共8页

摘　　要： 现有物联网系统普遍采用先来先服务的单一资源调度策略,不能很好地满足实时性和个性化需求.针对物联网系统的复杂性及需求目标的多样性,提出了一种多QoS驱动的物联网资源分层调度架构,建立了面向物联网的多QoS定义及形式化模型,并在此基础上提出了一种基于信息熵的、包含分类请求和精准选择两个过程的多QoS驱动的调度方法.仿真结果表明该资源调度方法能有效保障用户请求相对公平性,满足不同用户的实时性和个性化需求,提高了资源的利用率. Aimed to the Internet of Things （IoT） systems generally used a single resource scheduling strategy, which couldn＆#39;t meet the real-time and personalized requirements.To accommodate the complexity of IoT system and the diversity of the requests, the concept and formal model using a hierarchical scheduling architecture driven by multi-Quality of Service （QoS） resources was first proposed.A multi-QoS scheduling method was explored based on information entropy and contains request classification and precise selection processes.Simulation results show that the method can effectively ensure the relative fairness of user requests, meet the real-time and individual needs of different users, and also improve the resource utilization rate.

关 键 词： 分层调度 多 物联网