基于异构并行的DAS高密度数据实时解调技术

Real-time demodulation technology for high-density data in DAS based on heterogeneous parallelism

针对分布式光纤声波传感(Distributed optical fiber acoustic sensing,DAS)系统中高密度数据实时解调的需求,提出了基于中央处理器(Central processing unit,CPU)和图形处理器(Graphic processing unit,GPU)的异构并行计算架构,完成了实时解调双通道外差型DAS系统传感数据,可满足同时对两个通道共5000个等效阵元实时解调处理需求。此系统每秒需处理的数据量高达400 MB,相较于仅使用CPU运算的225.5 s运算时间,采用异构并行计算架构的运算时间优化到了468.2 ms,运算速度提升了482倍,且该方案仍有巨大的算力冗余空间,可为后续DAS系统整体实时性能的提升提供算力支持。

The heterogeneous parallel computing architecture based on Central processing unit(CPU)and Graphics processing unit(GPU)is proposed,to real-time demodulate the sensing data of a dual channel heterodyne DAS system in response to the demand for high-density data real-time demodulation in Distributed optical fiber acoustic sensing(DAS)systems.This architecture can meet the real-time demodulation and processing requirements of 5000 equivalent array elements in two channels simultaneously.The system needs to process up to 400 MB of data per second,which is 468.2 ms longer than the 225.5 s computation time using only CPU.The computation speed has been increased by 482 times using heterogeneous parallel computing architecture.This solution still has huge computing power redundancy space,which can provide computing power support for the overall real-time performance improvement of subsequent DAS systems.