Subjects: Astronomy >> Astrophysics submitted time 2023-01-06
Abstract:
"
"
"
"
Peer Review Status:Awaiting Review
Subjects: Astronomy >> Astrophysical processes submitted time 2022-06-28
Abstract:
After years of planning for the next generation of radio telescopes, the Square Kilometer Array (SKA), the construction of the SKA phase one (SKA1) had started in July 2021.After the formal operation of SKA1, it is expected that 750 petabytes of scientifically processed data will be generated every year. The data will be stored at SKA regional centers around the world for further analysis by researchers.In this paper, the models of SKA observation station, central signal processor, scientific data processing and regional center are quantitatively analyzed. Based on the high-priority scientific observation of SKA1, the data flow evaluation at each stage and the demand for computing power of scientific data processing are obtained. Taking the current SKA1-Low and SKA1-Mid arrays as examples, the key factors affecting the layout of interference arrays including resolution, sensitivity and UV coverage are summarized. Finally, OSKAR is used for data simulation of interference array. Through the simulation of SKA1-Mid, the scalability and stability of the system are obtained. Through the simulation of SKA1-Low on CSRC-P, it can be seen that the design of prototype SKA regional center in China has been fully optimized. And the detailed requirements of computing power and the detailed information of data volume are obtained. The SKA's demand for data processing, computing and storage also requires a combination of technologies and interdisciplinary efforts from areas such as electronics, communication, information technology and computer.
Peer Review Status:Awaiting Review
Subjects: Astronomy >> Astrophysical processes submitted time 2022-06-28
Abstract:
The Square Kilometer Array (SKA) is the largest radio telescope, and the data generated by its observations will be transmitted from Australia and South Africa to the scientific data processing center about one hundred kilometers away at first, and then distributed to various SKA Regional Centres(SRC) with a distance of tens of thousands of kilometers through high-speed network.In the SKA Phase One (SKA1) stage with a scale of 10\% of SKA, it is estimated that about 750PB of data needs to be distributed to each SRC through a network of at least 100Gbps each year. Such high network bandwidth and data scale bring great challenges to data transmission and distribution. This paper analyzes different network protocols such as TCP/UDP/HTTP and uses different software in the field of radio astronomy for testing and research, and then the optimal transmission scheme parameters under the current infrastructure of 10Gbps network are obtained. In this paper, the factors affecting high-speed transmission are discussed, and the corresponding performance optimization strategies are given.Before the real observation data of SKA1 is generated, it will provide the technical foundation for the network construction and layout of China's SKA regional center. The technical details and methods described are available for reference and use in relevant scientific applications. Finally, the challenges of future SKA network requirements are discussed and prospected.
Peer Review Status:Awaiting Review
Subjects: Astronomy >> Astronomical Instruments and Techniques submitted time 2022-06-28
Abstract:
We introduce a machine learning FRB dataset that can train the ML algorithms to reach the FRBs in raw data. It has 8020 FRB simulation images, 4010 non-FRB and 4010 RFI simulation images built from the public FRB observations, and can be expanded in any number as needed. This work provides an open-source dataset for state of art AI to the comparison of FRB event recognition algorithms. The dataset provides image and NumPy format files for both convolutional neural networks and classic machine learning algorithms. The dataset can implement FRB/non-FRB classification, or FRB/RFI/Blank classification. In the example, we used 31 pre-trained classic CNNs. In FRB/non-FRB classification, it achieves the accuracy of 90-92% in the first training epoch and max accuracy of 99.8% in real FRB dataset testing.
Peer Review Status:Awaiting Review