Unified Cramér-Rao bound for quantum multi-parameter estimation: Invertible and non-invertible Fisher information matrix

Abstract: In quantum multi-parameter estimation, the uncertainty in estimating unknown parameters is lower-bounded by Cram\'{e}r-Rao bound (CRB), defined as an inverse of the Fisher information matrix (FIM) associated with the multiple parameters. However, in particular estimation scenarios, the FIM is non-invertible due to redundancy in the parameter set, which depends on the probe state and measurement observable. Particularly, this has led to the use of a weaker form of the CRB to bound the estimation uncertainty in distributed quantum sensing. This weak CRB is generally lower than or equal to the exact CRB, and may, therefore, overestimate the achievable estimation precision. In this work, we propose an alternative approach, employing the Moore-Penrose pseudoinverse of the FIM for constrained parameters, providing a unified CRB, attainable with an unbiased estimator. This allows us to construct simple strategies for each case in both simultaneous estimation and distributed quantum sensing, covering paradigmatic examples considered in the literature. We believe this study to provide a unified framework for addressing non-invertible FIMs and improving the precision of quantum multi-parameter estimation in various practical scenarios.