基于MPI的卫星遥感数据实时处理平台设计

Design of Real-time Processing Platform for Satellite Remote Sensing Data Based on MPI

针对高速率卫星遥感数据实时处理的要求,设计了一种卫星遥感数据实时处理平台。该平台采用分层设计思想,自上而下分为表现与控制层、业务处理层、基础服务层和应用支持层。结合消息传递接口(MPI)技术,构建了分布式实时并行处理平台框架。原型系统验证结果表明,该平台能处理下传码速率达450Mbit/s的双通道数据,单通道处理速率高达6Gbit/s。文章提出的平台设计应用于遥感卫星的地面检测系统,具有处理速度快、运行稳定的优点。

A platform is designed for high speed and real-time processing of satellite remote sensing data.The platform takes advantage of layered architecture system which,from top to bottom,involves presentation and control layer,business treatment layer,basic service layer,and application support layer.With MPI（message passing interface）technology,a framework is built for distributed real-time parallel processing and demonstrated by aprototype system.The results show that the platform is able to handle double channel data with a download code rate up to 450Mbit/s,and single channel data with processing rate above 6Gbit/s.The platform is applied to the ground test system of remote sensing satellites,and has the advantages of high processing speed and stable operation.