Multichannel infrared receiver performance

The performance of an optical receiver designed to detect a target by taking advantage of the target's spectral signature is presented. The receiver processes the signal in several narrow frequency bands and is based upon a statistical model which represents the field in each band as a Gaussian random process whose moments depend upon the target and background characteristics. The optimal Bayes/Neyman-Pearson receiver structure for an M spectral channel, N sequential look receiver is presented, and practical suboptimal receiver structures are developed. Numerical methods are used to calculate the probability of false alarm and the probability of detection using identical parameters for each processor. The results indicate that the approximate receiver structures and the ad hoc receiver structures all have the same performance. The results also show that performance depends only upon the difference in the square root of the mean to variance ratios under each hypothesis and the ratio of the variances.