【24h】

MITRE Sensor Layer Prototype

机译:MITER传感器层原型

获取原文
获取原文并翻译 | 示例

摘要

The MITRE Sensor Layer Prototype is an initial design effort to enable every sensor to help create new capabilities through collaborative data sharing. By making both upstream (raw) and downstream (processed) sensor data visible, users can access the specific level, type, and quantities of data needed to create new data products that were never anticipated by the original designers of the individual sensors. The major characteristic that sets sensor data services apart from typical enterprise services is the volume (on the order of multiple terabytes) of raw data that can be generated by most sensors. Traditional tightly coupled processing approaches extract pre-determined information from the incoming raw sensor data, format it, and send it to predetermined users. The community is rapidly reaching the conclusion that tightly coupled sensor processing loses too much potentially critical information.' Hence upstream (raw and partially processed) data must be extracted, rapidly archived, and advertised to the enterprise for unanticipated uses. The authors believe layered sensing net-centric integration can be achieved through a standardize-encapsulate-syndicate-aggregate-manipulatc-process paradigm. The Sensor Layer Prototype's technical approach focuses on implementing this proof of concept framework to make sensor data visible, accessible and useful to the enterprise. To achieve this, a "raw" data tap between physical transducers associated with sensor arrays and the embedded sensor signal processing hardware and software has been exploited. Second, we encapsulate and expose both raw and partially processed data to the enteiprise within the context of a service-oriented architecture. Third, we advertise the presence of multiple types, and multiple layers of data through geographic-enabled Really Simple Syndication (GeoRSS) services. These GeoRSS feeds are aggregated, manipulated, and filtered by a feed aggregator. After filtering these feeds to bring just the type and location of data sought by multiple processes to the attention of each processing station, just that specifically sought data is downloaded to each process application. The Sensor Layer Prototype participated in a proof-of-concept demonstration in April 2008. This event allowed multiple MITRE innovation programs to interact among themselves to demonstrate the ability to couple value-adding but previously unanticipated users to the enterprise. For this event, the Sensor Layer Prototype was used to show data entering the environment in real time. Multiple data types were encapsulated and added to the database via the Sensor Layer Prototype, specifically National Imagery Transmission Format 2.1 (NITF), NATO Standardization Format 4607 (STANAG 4607), Cursor-on-Target (CoT), Joint Photographic Experts Group (JPEG), Hierarchical Data Format (HDF5) and several additional sensor file formats describing multiple sensors addressing a common scenario.
机译:MITER传感器层原型是一项初始设计工作,旨在使每个传感器都可以通过协作数据共享来帮助创建新功能。通过使上游(原始)和下游(已处理)传感器数据均可见,用户可以访问特定级别,类型和数量的数据,以创建单个传感器的原始设计者从未期望的新数据产品。将传感器数据服务与典型企业服务区分开来的主要特征是,大多数传感器可以生成的原始数据量(数TB级)。传统的紧密耦合处理方法从传入的原始传感器数据中提取预定信息,对其进行格式化,然后将其发送给预定用户。社区迅速得出结论,紧密耦合的传感器处理会丢失太多潜在的关键信息。因此,上游(原始和部分处理的)数据必须被提取,快速存档并发布给企业,以备不时之需。作者认为,分层感应网络中心集成可以通过标准化-封装-辛迪加-聚集-操纵过程范例来实现。传感器层原型的技术方法侧重于实施此概念证明框架,以使传感器数据对企业可见,可访问且有用。为了实现这一点,已经利用了与传感器阵列相关联的物理换能器与嵌入式传感器信号处理硬件和软件之间的“原始”数据抽头。其次,我们在面向服务的体系结构的上下文中封装原始数据和部分处理的数据,并将它们公开给企业。第三,我们通过启用地理位置的Really Simple Syndication(GeoRSS)服务来宣传多种类型和多层数据的存在。这些GeoRSS提要由提要聚合器聚合,操纵和过滤。在对这些提要进行过滤以使多个处理程序仅寻求各处理站关注的数据的类型和位置之后,将专门寻找的数据下载到每个处理程序应用程序中。 Sensor Layer Prototype参加了2008年4月的概念验证演示。此活动使多个MITER创新计划可以相互进行交互,以展示将增值但先前未曾想到的用户耦合到企业中的能力。对于此事件,使用了“传感器层原型”来实时显示进入环境的数据。封装了多种数据类型,并通过传感器层原型将其添加到数据库中,特别是国家图像传输格式2.1(NITF),北约标准化格式4607(STANAG 4607),目标光标(CoT),联合图像专家组(JPEG) ),分层数据格式(HDF5)和其他几种传感器文件格式,这些文件描述了解决常见情况的多个传感器。

著录项

相似文献

  • 外文文献
  • 中文文献
  • 专利
获取原文

客服邮箱:kefu@zhangqiaokeyan.com

京公网安备:11010802029741号 ICP备案号:京ICP备15016152号-6 六维联合信息科技 (北京) 有限公司©版权所有
  • 客服微信

  • 服务号