The Research Of Methods To Search Fast Radio Bursts

Radio transient plays an important role in Radio Astronomy.Since the first Pulsar was found in 1960s,nearly 3000 pulsars have been detected.The Nobel Pr:ice was won twice for the great contribution of the discovery of pulsar and binary pulsar system separately.Fast Radio Burst(FRB)is also an important radio transient.Published in 2007,Lorimer et al.found the first fast radio burst in archive pulsar data with the Australian 64-meter Parks radio telescope.Compared to the pulses emitted by pulsars,FRB does not have periodic pulsed radiation,but its energy is hundreds of times higher than the pulsar pulse.Nearly 100 cases of FRB have been discovered,and most of the FRBs only occur once,but two cases of FRBs have been repeatedly detected which caused a great sensation in the astronomical world.The FRB 121102 is one of repeated FRBs.It was first discovered by the Arecibo telescope in the United States.The Very Large Array(VLA)was followed observing in 2016,due to the advantage of long baseline,the source position is accurate to 1.7 mili-arcsecond.In the early 2019,CHIME team announced the second repeated FRB that caused widespread concern.This thesis also details the construction methods and design techniques of the transient search digital backend.Detailed research on methods to search FRB were done from mathematic De-dispersion algorithm to the radio telescope FRB search principle and from hardware back-end to software data processing.We also introduce the Collaboration for Astronomy Signal Processing and Electronics Research(CASPER)hardware platform.Using this platform can greatly save FRB back-end design time.Combined with my research projects during my PhD program,we presented some examples of FRB search backend design.The design skills and techniques are summarized.Different digital backend design solutions are present for the specific needs of different radio telescopes.In this thesis,a method for searching radio transient is proposed.We search radio transients in two-dimensional Fourier(2D-FFT)space.This method can effectively reduce search time and improve search efficiency and sensitivity.We also provide a search software based on the search algorithm mentioned above,which we call it 2DFFT.This radio transient search software is written in Python and can be parallelized easily.Due to good parallel ability,this algorithm can be easily ported to the GPU in the future work,and it is expected the performance will improve dramatically at that time.