Parallel Optimization Of W-term Correction Algorithm Based On DALiuGE Computing Framework

The Square Kilometer Array(SKA)will be the largest integrated aperture radio telescope ever built in the world.The SKA is the next generation of radio telescope with the core concepts of large field of view,multiple beams,high dynamics,high resolution,and big data.It will revolutionize the traditional research methods of radio astronomy and bring revolutionary and brand-new ideas to astronomy research.With the proposal of the SKA program,radio interference array imaging is currently a research hotspot.This paper analyzes the overall structural design of SKA,the acquisition of observational data and the process of sampling and imaging,and the characteristics of real-time processing of SKA massive observation data.It is based on a distributed computing framework DALiuGE w-term correction algorithm adopted by SKA.Achieved research to meet the SKA follow-up experimental work.The details are as follows:(1)Facing the distributed processing requirements of SKA Scientific Big Data,the data acquisition and processing flow of the radio interference array was analyzed and investigated.The principle of comprehensive aperture imaging and large-field imaging were carefully studied,and the development method of the current DALiu GE computing framework was carried out.Learning and experimentation.(2)Focusing on the unique requirements of SKA with large field of view,high dynamics,and high resolution imaging,the key technologies of large-field radio interference imaging are researched.The faceting algorithm is studied in depth,and the algorithm is implemented and tested initially.(3)On the basis of work two,according to the requirements of the SKA scientific data processor work package(SDP),the Faceting algorithm is further packaged as a DROP and implemented under the DALiuGE framework.In general,this paper studies and implements the parallel optimization of the w correction algorithm based on the DALiuGE computational framework,which lays a good foundation for the SKA imaging algorithm and makes innovations in the implementation of the high-performance massive data processing pipeline.It provides reference and reference for SKA observation and data processing.