Uncertainty analysis in the measurement of the spatial responsivity of infrared antennas.

The measurement of a two-dimensional spatial responsivity map of infrared antennas can be accomplished by use of an iterative deconvolution algorithm. The inputs of this algorithm are the spatial distribution of the laser beam irradiance illuminating the antenna-coupled detector and a map of the measured detector response as it moves through the illuminating beam. The beam irradiance distribution is obtained from knife-edge measurements of the beam waist region; this data set is fitted to a model of the beam. The uncertainties, errors, and artifacts of the measurement procedure are analyzed by principal-component analysis. This study has made it possible to refine the measurement protocol and to identify, classify, and filter undesirable sources of noise. The iterative deconvolution algorithm stops when a well-defined threshold is reached. Spatial maps of mean values and uncertainties have been obtained for the beam irradiance distribution, the scanned spatial response data, and the resultant spatial responsivity of the infrared antenna. Signal-to-noise ratios have been defined and compared, and the beam irradiance distribution characterization has been identified as the statistically weakest part of the measurement procedure.