Information Theoretic Limits for Phase Retrieval With Subsampled Haar Sensing Matrices
Information Theoretic Limits for Phase Retrieval With Subsampled Haar Sensing Matrices
复制标题
子采样 Haar 传感矩阵相位检索的信息论极限
DOI:
10.1109/tit.2020.3015173
复制
发表时间:
2019
影响因子:
2.5
通讯作者:
A. Maleki
中科院分区:
文献类型:
--
作者:
Rishabh Dudeja;Junjie Ma;A. Maleki
We study information theoretic limits of recovering an unknown <inline-formula> <tex-math notation="LaTeX">${n}$ </tex-math></inline-formula> dimensional, complex signal vector <inline-formula> <tex-math notation="LaTeX">${x}_{\star} $ </tex-math></inline-formula> with unit norm from <inline-formula> <tex-math notation="LaTeX">${m}$ </tex-math></inline-formula> magnitude-only measurements of the form <inline-formula> <tex-math notation="LaTeX">${y}_{i} = |({A}{x}_{\star} )_{i}|^{2}, \; {i} = 1,2 {\dots }, {m}$ </tex-math></inline-formula>, where <inline-formula> <tex-math notation="LaTeX">${A}$ </tex-math></inline-formula> is the sensing matrix. This is known as the Phase Retrieval problem and models practical imaging systems where measuring the phase of the observations is difficult. Since in a number of applications, the sensing matrix has orthogonal columns, we model the sensing matrix as a subsampled Haar matrix formed by picking <inline-formula> <tex-math notation="LaTeX">${n}$ </tex-math></inline-formula> columns of a uniformly random <inline-formula> <tex-math notation="LaTeX">${m} \times {m}$ </tex-math></inline-formula> unitary matrix. We study this problem in the high dimensional asymptotic regime, where <inline-formula> <tex-math notation="LaTeX">${m},{n} \rightarrow \infty $ </tex-math></inline-formula>, while <inline-formula> <tex-math notation="LaTeX">${m}/{n} \rightarrow \delta $ </tex-math></inline-formula> with <inline-formula> <tex-math notation="LaTeX">$\delta $ </tex-math></inline-formula> being a fixed number, and show that if <inline-formula> <tex-math notation="LaTeX">${m} < (2-{o}_{n}(1))\cdot {n}$ </tex-math></inline-formula>, then <italic>any estimator</italic> is asymptotically orthogonal to the true signal vector <inline-formula> <tex-math notation="LaTeX">${x}_{\star} $ </tex-math></inline-formula>. This lower bound is sharp since when <inline-formula> <tex-math notation="LaTeX">${m} > (2+{o}_{n}(1)) \cdot {n} $ </tex-math></inline-formula>, estimators that achieve a non trivial asymptotic correlation with the signal vector are known from previous works.
登录
查看更多内容
影响因子:
5.4
作者:
Luo, Wangyu;Alghamdi, Wael;Lu, Yue M.
通讯作者:
Lu, Yue M.
影响因子:
2.5
作者:
Goldstein, Tom;Studer, Christoph
通讯作者:
Studer, Christoph
DOI:
--
发表时间:
2019
期刊:
Proceedings of Machine Learning Research
影响因子:
--
作者:
Reeves, Galen;Xu, Jiaming;Zadik, Ilias
通讯作者:
Zadik, Ilias
影响因子:
3
作者:
Mondelli, Marco;Montanari, Andrea
通讯作者:
Montanari, Andrea