Multi-criteria decision algorithm for NQR signal detection

Nuclear quadrupole resonance is a spectroscopy technique used for chemical analysis, temperature measurement and detection of prohibited substances. It offers advantages like highly specific and non-invasive detection, but has the disadvantages of low signal-to-noise ratio and temperature dependence of the response signal, which are generally compensated by special hardware and signal processing solutions. This paper proposes a data processing algorithm based on multiple criteria designed to ensure nuclear quadrupole resonance detection in specific scenarios. In contrast with current algorithms, this solution does not require statistical estimation of the response's parameters. It is evaluated using a real data set composed of 8000 acquisitions and is shown to reach a detection probability of 94% that outperforms existing solutions for a specific detection threshold.