I. Introduction
Signal and subspace detection in the presence of disturbance is an important topic in various areas such as medical imaging [1], [2], radar [3], communication systems [4] and pattern recognition [5]. Neyman and Pearson [6] showed that the likelihood ratio test (LRT) is the uniformly most powerful (UMP) test when the distributions of the data under both hypotheses and are known. Unfortunately, the UMP test does not exist in many signal processing applications due to uncertainties and the presence of unknown parameters in the assumed models [7]. Thus, finding suboptimal detectors is always of interest to researchers from different background to achieve better detection results in their targeted applications [8].