Answers are grounded in this paper and cite their sources.
Quickest change detection (QCD) is a fundamental statistical framework for identifying abrupt anomalies in streaming data. Translating this framework to the quantum domain introduces new challenges. In this paper, we propose a unified optimization framework for QCD of quantum states and quantum channels under varying degrees of prior knowledge. Specifically, for QCD of quantum states, when the post-change state is known, we show that the optimal positive operator-valued measure (POVM) that maximizes the measured Kullback–Leibler (KL) divergence between pre- and post-change states is von Neumann, and develop a projected gradient ascent algorithm that computes a locally optimal von Neumann POVM. When the post-change state is unknown, we show that the maximum-sensitivity POVM is given by the eigen-projectors corresponding to the positive eigenvalues of the pre-change state, and propose a window-based CUSUM-like test. We show that the rank-deficient pre-change state substantially reduces the detection delay via null-space projection. Based on this property, for QCD of quantum channels, we jointly optimize the probe state and the POVM, with the primary objective of inducing rank-deficiency of the channel output state, and the secondary objective of maximizing the KL divergence (for known post-change channel) or the sensitivity (for unknown post-change channel).
No takes yet. Share an insight, caveat, or question.
Fu et al. (2026) studied this question.
Synapse has enriched 10 related papers. Consider them for comparative context: