NQR signal processing based on multi-stage Wiener filter

Nuclear quadrupole resonance (NQR) offers an unequivocal method of detecting and identifying explosives. Unfortunately, the inherent signalto-noise of the most Ammonium Nitrate (AN) is very low, For the purpose of detecting weak NQR echo train from signal which is corrupted by RF interference (RFI) and noise, a NQR signal processing method based on multi-stage wiener filter (MSWF) is presented. MSWF, implementing the wiener filter using a nested chain of scalar wiener filters, is a reduced-rank algorithm. Compared with the standard method of solving Wiener filter, MSWF does not need matrix inversion, effectively reduces the computational complexity. In this work, we employ ammonium nitrate (AN) with different weight as the experimental samples; treat the echo train as reference channel in the condition of no sample, and other conditions as observed channel. Processing steps are as follows, applying MSWF to eliminate noise and RFI by utilizing the correlation between reference channel and observed channel, accumulating a certain times and then using the fast Fourier transformation to transform the signal to frequency domain, clear spectrum peak is achieved at the resonance frequency of AN. The results indicate that the method eliminates RFI and noise efficiently, obtains stable detection ability, reduces the computational complexity effectively, and makes the real-time NQR signal detection to be possible.