传感器网络中协作任务的实时调度  

Real-time Scheduling for Cooperative Tasks in Sensor Networks

作　　者：胡侃[1] 刘云生[1] 

机构地区：[1]华中科技大学计算机学院,武汉430074

出　　处：《计算机科学》2007年第10期65-69,共5页Computer Science

基　　金：国家自然科学基金项目(60073045);博士点基金及国防预研基金(00J15.3.3.JW0529)资助

摘　　要：在传感器网络实时监测应用中,大量传感器散布在监测区域中感知监测域的各种环境或监测对象的信息,一组功能有限的传感器往往相互协作地完成一个大的实时感知任务,协作性是传感器网络的重要特性,它要求实时任务之间的资源共享。单纯的实时系统为保证任务的实时性通常采用资源隔离机制而不能很好地解决传感器网络环境中的采集流数据处理的协作性问题。本文基于服务器的调度框架,使用了实时环境中的时间属性,将数据时间与程序时间相结合,从而将应用语义与系统中运行的程序相联系,提出了一种基于时间依赖关系的实时调度模式,并给出了基于此模式的事件驱动并发数据流程图模型及其实现机制。分析表明,该模型能有效地解决传感器网络监测区域中采集流数据处理过程的协作性问题,减少了数据丢失,提高了系统响应的实时性。Wireless sensor networks are being developed for a variety of applications. With the continuing advances in network and application design, appropriate middleware is needed to provide the capability to support efficient real-time QoS for coordinative concurrent applications on sensor networks. In traditional simple real-time system, resource isola- tion mechanism is utilized for ensuring the real-time response of the tasks, it is futile to the problems where streaming data are processed collaboratively by concurrent fusion tasks and timely responses are important, The paper studies a schedule approach for providing efficient real-time services to concurrent coordinative tasks in sensor networks. We first propose a coordinative execution scheme for data stream processing. And then, by evolving server scheduling techniques and combining timestamp of data stream with the task execution time to tailor the design of schedule software, we present a three-layer scheduling architecture to provide real-time services while guaranteeing coordination consistency. Finally, we design a three-phrase scheduling algorithms based on the model. Our performance evaluations show that by applying the three-phrase scheduling algorithms, the outer transformation ratio of collection data can be significantly enhanced and the amount of lost data can be reduced correspondingly.

关 键 词：实时系统 协作执行 流数据 数据流模型 传感器网络 

分 类 号：TP212[自动化与计算机技术—检测技术与自动化装置]