Track-before-detect for sensors with complex measurements

Abstract

Track before detect (TkBD) is a paradigm that combines the target detection and estimation processes that are usually sequentially applied to sensor data in a conventional system. Under TkBD the single frame detector is removed and the tracker is supplied with the whole sensor image. Detection decisions are then shifted to the output of the tracker which is able to use temporal correlation to improve the decision performance. A fundamental measure used by most TkBD approaches is the likelihood ratio of the sensor data and this is formed as the product over individual cell likelihoods under the assumption of spatially independent noise. However, that approach exploits only the envelope of the known sensor point spread function. This article presents an approach for determining the data likelihood ratio that also includes phase information. This alternative likelihood ratio formulation is shown to both improve the discrimination of targets from noise and reduce the computation overhead of the algorithm.