面向物联网的工作流服务架构模型  

IOT-ORIENTED WORKFLOW SERVICE ARCHITECTURE MODEL

作　　者：罗国涛[1] 裴广[2] 邓祖朴 刘勇军[1] 

机构地区：[1]四川托普信息技术职业学院计算机学院,四川成都611743 [2]中国科学院成都计算机应用研究所图像部,四川成都610041

出　　处：《计算机应用与软件》2014年第2期9-13,共5页Computer Applications and Software

摘　　要：为有效解决工作流系统与物联网的协同工作问题,以及针对以往工作流系统在分布式架构和海量数据处理中的不足,提出面向物联网的工作流服务架构模型。该模型将工作流表单、数据库服务以及工作流任务等封装成组件,并将组件以可插拔的方式注入到工作流服务集控中心容器中,从而实现物联网模式下资源的即插即用以及所需即所得;RFID中间件集成Jini服务使得工作流系统很好地与RFID通信,提高了工作流处理表单数据的效率;在底层海量数据处理上,提出一种改进的MapReduce模型,该模型在HDFS的架构方案基础上采用任务节点元数据定时备份策略,解决了任务节点失效后所带来的系统运行不稳定问题。最后对工作流服务模型中各关键要素提出整合方案,并通过实验证明了该方案的有效性。To solve the collaboratively working between workflow system and IOT effectively,and aiming at the deficiencies of previous workflow systems in distributed architecture and massive data processing,we propose an IOT-oriented workflow service architecture model. The model encapsulates the workflow worksheet,database service and workflow tasks into components and infuses them into workflow service control centre container in a pluggable way,so as to achieve the plug-and-play and the what-required-is-what-gotten for the resource in IOT mode.RFID middleware integrating the Jini service enables the communication between the workflow system and the RFID works well,this increases the efficiency of worksheet data processing of the workflow.In massive bottom layer data processing,we propose an improved MapReduce model.It adopts task-node metadata timing backup strategy based on the scheme of HDFS architecture,and solves the problem of unstable system running caused by task-node failure.In the end,we put forward an integration programme for various key elements in work-flow service model,and prove its effectiveness through the experiments.

关 键 词：工作流服务组件 物联网 RFID中间件 Jini服务 元数据 Internet of THING (IOT) 

分 类 号：TP311.52[自动化与计算机技术—计算机软件与理论]