RFI MITIGATION IMPLEMENTATION FOR PULSAR RADIOASTRONOMY D AITALLAL
- Slides: 16
RFI MITIGATION IMPLEMENTATION FOR PULSAR RADIOASTRONOMY D. AITALLAL, C. DUMEZ-VIOU, R. WEBER Observatoire de Paris /CNRS-INSU CNRS B 704 www. obs-nancay. fr Project Prep. SKA (contract no 212243) and Project Radionet/Uniboard (contract no 227290)), The French fundingagency ANR (contract ANR-09 -BLAN-0225 -04)
RFI MITIGATION IMPLEMENTATION FOR PULSAR RADIOASTRONOMY Outline: I. Introduction II. Power Detector at Nançay multipurpose receiver (RDH) III. Cyclo. Det: A cyclic detector. IV. Two Dimensional FFT detector (2 D-FFT): RFI mitigation + Giant Pulses detection V. Conclusion 2
Introduction For Pulsar Timing: a) impulsive or burst broad-band RFI => Power criteria Detector b) Just after the PFB, for continuous RFI Cyclostationary c) Just after the FFT, for narrow band continuous RFI Detector For Giant Pulse detection: On-line method detection with RFI mitigation capabilities both for impulsive and narrow band RFI. Based on a 2 -dimensional FFT (2 D FFT) and Radon transform. 3
RFI MITIGATION IMPLEMENTATION FOR PULSAR RADIOASTRONOMY Outline: I. Introduction II. Power Detector at Nançay multipurpose receiver (RDH) III. Cyclo. Det: A cyclic detector. IV. Two Dimensional FFT detector (2 D-FFT): RFI mitigation + Giant Pulses detection V. Conclusion 4
Power Detector Decimeter Radiotelescope Decameter array • High dynamic (70 d. B) • High frequency resolution (Δfmin=107 Hz) • High temporal resolution (Δtmin=9 µs/spectrum) • FPGA and DSP resources available for RFI mitigation 5
Blanker on waveform (1) Architecture optimized for efficient hardware implementation : Shift registers, comparators, multipliers, glue logic 2. 7% of a XC 2 V 3000 FPGA 145 Ms/s throughput Robust mean power estimation Strong pulses strategy 3 samples window ~3σ detection Weak pulses strategy 30 samples window ~0. 8σ detection Blanking decision logic 6
Blanker on waveform (2) Testing of the radar blanker on real data in real time RADAR PGC 51094 with blanking without blanking 7
RFI MITIGATION IMPLEMENTATION FOR PULSAR RADIOASTRONOMY Outline: I. Introduction II. Power Detector at Nançay multipurpose receiver (RDH) III. Cyclo. Det: A cyclic detector IV. Two Dimensional FFT detector (2 D-FFT): RFI mitigation + Giant Pulses detection V. Conclusion 8
Cyclostationarity concept Cyclostationary process Stationary process statistics time-independent Example : second order statistics s(t) or n(t) statistics are periodic T = hidden periodicity r(t) Cyclic signature 0 Cyclic correlation Cyclic frequency In practice : 9 Gardner and Giannakis 0 Cyclic frequency
CYCLODET: A CYCLIC DETECTOR 10
RFI MITIGATION IMPLEMENTATION FOR PULSAR RADIOASTRONOMY Outline: I. Introduction II. Power Detector at Nançay multipurpose receiver (RDH) III. Cyclo. Det: A cyclic detector IV. Two Dimensional FFT detector (2 D-FFT): RFI mitigation + Giant Pulses detection V. Conclusion 11
ON-line Giant Pulse Detector 12
Giant Pulse Detector 2 D-FFT Radon Transform Without blanking 13 With blanking
Giant Pulse Detector 14
RFI MITIGATION IMPLEMENTATION FOR PULSAR RADIOASTRONOMY Outline: I. Introduction II. Power Detector at Nançay multipurpose receiver (RDH) III. Cyclo. Det: A cyclic detector IV. Two Dimensional FFT detector (2 D-FFT): RFI mitigation + Giant Pulses detection V. Conclusion 15
Conclusion • In the framework of the UNIBOARD FP 7 European project, these algorithms will be implemented in a multi-purpose scalable computing platform for Radio Astronomy as part of the pulsar receiver. • In the case of Giant pulse detection, a new approach which combines a hardware efficient search method and some RFI mitigation capabilities has been proposed. • Extend this 2 D-FFT tool for pulsar search 16