Low-Complexity GLRT Based Quickest Detection With Unknown Parameters

Consider a quickest detection problem, where a sudden change of parameters needs to be detected as quickly as possible. When unknown parameters exist in the post-change distribution, generalized likelihood ratio test (GLRT) based method is a straightforward option, which may, however, lead to huge computational complexity and storage burden due to the repeated operation of the enumerating all possible change time and estimating unknown parameters every time a new observation sample is received. In this paper, a drift-oriented GLRT (D-GLRT) algorithm is proposed to avoid the repeated enumerations and reduce the storage burden. It is shown that the D-GLRT has lower complexity compared with the conventional GLRT based method, with a little loss of performance. The upper bound of the worst case average detection delay (WADD) of the D-GLRT based method is derived. Numerical results are provided to validate our theoretical analysis.