https://hal-utt.archives-ouvertes.fr/hal-02297226Nikiforov, IgorIgorNikiforovLM2S - Laboratoire Modélisation et Sûreté des Systèmes - ICD - Institut Charles Delaunay - UTT - Université de Technologie de Troyes - CNRS - Centre National de la Recherche ScientifiqueA suboptimal quadratic change detection schemeHAL CCSD2000[SPI.SIGNAL] Engineering Sciences [physics]/Signal and Image processingVU VAN, Jean-Baptiste2019-09-25 18:30:552023-03-24 14:53:122019-09-25 18:30:55enJournal articles10.1109/18.8684801We address the problem of detecting changes in multivariate Gaussian random signals with an unknown mean after the change. The window-limited generalized-likelihood ratio (GLR) scheme is a well-known approach to solve this problem. However, this algorithm involves at least (log /spl gamma/)//spl rho/ likelihood-ratio computations at each stage, where /spl gamma/(/spl gamma//spl rarr//spl infin/) is the mean time before a false alarm and /spl rho/ is the Kullback-Leibler information. We establish a new suboptimal recursive approach which is based on a collection of L parallel recursive /spl chi//sup 2/ tests instead of the window-limited GLR scheme. This new approach involves only a fixed number L of likelihood-ratio computations at each stage for any combinations of /spl gamma/ and /spl rho/. By choosing an acceptable value of nonoptimality, the designer can easily find a tradeoff between the complexity of the quadratic change detection algorithm and its efficiency.