CSR: EAGER: Exploratory Research on Data Quality-Centric Cloud Infrastructure for Federated Sensor Services

这个迫切的项目支持在传感器网络环境下探索数据质量(DQ)问题的基础研究。具体来说，这个项目将包括三个主要的研究项目：(1)能够适应AdiVerse的DQ属性集的传感器馈送的多维DQ模型。该项目将对各种DQ维度进行全面研究，包括发明用于准确量化DQ维度的指标和机制；(2)该项目将为联合传感器服务云创建面向服务的架构框架，其中DQ感知渗透到架构的所有级别和所有方面。我们的核心思想是以DQ感知的方式虚拟化所有传感器和辅助服务；以及(3)该项目将设计DQ感知的自主技术，用于传感器服务发现、Feed组合和适应环境动态。我们的技术将通过检测在多个用户请求之间共享的提要处理任务，以及通过云拓扑感知提要处理任务的分布来增强可伸缩性。

This EAGER project supports foundational research toward the exploration of data quality (DQ) issues in the context of sensor networks. Specifically, this project will include three major research thrusts: (1) a multidimensional DQ model for sensor feeds that can accommodate adiverse set of DQ attributes. The project will carry out a comprehensive study into various DQ dimensions which will include inventing metrics and mechanisms for accurately quantifying the DQ dimensions; (2) The project will create a service oriented architectural framework for federated sensor service clouds in which DQ awareness is pervasive at all levels and all aspects of the architecture. Our core idea is to virtualize all sensors and auxiliary services in a DQ-aware manner; and (3) The project will devise DQ-aware autonomic techniques for sensor service discovery, feed composition and adaptation towards environmental dynamics. Our techniques will enhance scalability by detecting feed processing tasks that are shared among multiple user requests, as well as by cloud topology-aware distribution of feed processing tasks.