Research On The Frame Technology Of 40-meter Pulsar Data Processing In The PTZ Based On Stream Computing

The pulsar is a neutron star that rotates rapidly and continuously emits a pulse signal.Due to its special physical properties and extremely stable periodic changes,it can be used to study high-precision timing or plasma physics.Therefore,the related research on pulsars have great significance in modern astronomy.In recent years,research on new topics such as gravitational waves and navigation based on pulsars has placed higher demands on the performance of pulsar observation equipment.At present,two kinds of processing methods are used in the signal receiving and processing of pulsar observation abroad.One is based on real-time acquisition and processing.However,the overall framework is not flexible enough,there are many difficulties in the later development process;the other is based on offline file reading for signal acquisition and processing,but it can not meet the real-time requirements.In the research of pulsar observation data processing in China,due to the late start,with the introduction and use of high-throughput radio telescope observation equipment,most of them are based on offline file method or in the exploration stage of real-time processing method.Therefore,the high throughput back-end pulsar observation equipment urgently needs the introduction of corresponding data stream processing technology.Although the performance of computer hardware is developing rapidly,it is still difficult and timeconsuming to develop and maintain real-time streaming data processing technology,which seriously affects the scientific output.In order to solve the problem of data processing efficiency of pulsars in China,this paper mainly studies the real-time processing framework technology of flow data at home and abroad based on the way of collecting and post-processing pulsars in the process of data processing of 40-meter radio telescope at Yunnan Observatory,and analyses and compares the real-time processing framework of radio astronomical pulsar data and other related technologies.Based on the analysis of the results,a PulsarPipeline Pulsar Data Processing Framework is designed.The framework consists of three main components: data processing pipeline,functional function of algorithm and circular buffer connecting functional function.It realizes the streaming processing of pulsar data,and then realizes the framework model.On this basis,the various functional components of the decomposition algorithm for ROACH2 observation data are realized.Finally,a prototype of Pipeline flow processing framework for the decomposition processing of pulsar signals of Yunnan Observatory 40-meter radio telescope is constructed and verified.The experimental results show that the PulsarPipeline framework has certain performance advantages over the existing processing methods in the processing speed of the pulsar data.My research results and innovations are as follows:1)Based on the existing data processing of the 40-meter radio telescope at Yunnan Observatory,the design idea of PulsarPipeline Pulsar data processing framework is conceived.According to the different languages of the framework,the design is divided into user modeling layer design and internal computing engine design,and the detailed design description is given.2)Based on the PulsarPipeline framework,the pulsar data processing and the chromatic dispersion algorithm are implemented on the CPU.The input number is J0835-4510 pulsar observation data,and the contour map of the pulsar after the chromatic dispersion is obtained.The simulation method of the pulsar data processing of the Yunnan Observatory is performed.And the achromatic dispersion algorithm based on PulsarPipeline flow computing framework compares the efficiency of different sizes of data.3)The efficiency of the pulsar achromatic dispersion algorithm implemented on the PulsarPipeline stream processing framework is optimized,that is,the GPU-based pulsar data processing or achromatic chromatic algorithm is implemented,and the efficiency of the CPU achromatic dispersion algorithm and the GPU achromatic dispersion algorithm based on the PulsarPipeline flow calculation framework are performed.Comparative analysis demonstrates the advantages of the PulsarPipeline data processing framework.At the same time,the existing problems and future work will be discussed.