期刊
IEEE TRANSACTIONS ON SIGNAL PROCESSING
卷 71, 期 -, 页码 2391-2406出版社
IEEE-INST ELECTRICAL ELECTRONICS ENGINEERS INC
DOI: 10.1109/TSP.2023.3283047
关键词
Greedy sensor selection; Fisher information intensity; change of basis; volume ratio; submodularity
This article proposes greedy approaches to select informative sensors to maximize the Fisher information, and introduces a new metric called the Fisher information intensity (FII). The volume ratio between the information ellipsoid corresponding to the selected subset and the ground set is optimized. A cost function based on the volume ratio is developed and proven to be monotone submodular. A greedy algorithm and its fast version are proposed to obtain near-optimal solutions. Numerical results demonstrate the superiority of the proposed algorithms.
This article focuses on greedy approaches to select the most informative k sensors from N candidates to maximize the Fisher information, i.e., the determinant of the Fisher information matrix (FIM), which indicates the volume of the information ellipsoid (VIE) constructed by the FIM. However, it is a critical issue for conventional greedy approaches to quantify the Fisher information properly when the FIM of the selected subset is rank-deficient in the first (n - 1) steps, where n is the problem dimension. In this work, we propose a new metric, i.e., the Fisher information intensity (FII), to quantify the Fisher information contained in the subset S with respect to that in the ground set N specifically in the subspace spanned by the vectors associated with S. Based on the FII, we propose to optimize the ratio between VIEs corresponding to S and N. This volume ratio is composed of a nonzero (i.e., the FII) and a zero part. Moreover, the volume ratio can be easily calculated based on a change of basis. A cost function is developed based on the volume ratio and proven monotone submodular. A greedy algorithm and its fast version are proposed accordingly to guarantee a near-optimal solution with a complexity of O(Nkn(3)) and O(Nkn(2)), respectively. Numerical results demonstrate the superiority of the proposed algorithms under various measurement settings.
作者
我是这篇论文的作者
点击您的名字以认领此论文并将其添加到您的个人资料中。
推荐
暂无数据