Rank-One Signal Recovery in Sparse Wishart Noise
Abstract
We study the high-dimensional recovery of a signal vector x in the presence of sparse Wishart-like noise. We define an N × N matrix A = J+(θ/N)xx⊤, where xx⊤ is the rank-one deformation of the random noise matrix J. We consider a Wishart-like matrix J=X⊤ X, where X is a sparse M × N random matrix with entries Xij = cijWij, with cij regulating the density of non-zero elements, and Wij the bond weights. Using the replica method, we compute analytically the top eigenpair statistics of A, and their dependence on the signal strength θ, the rectangularity ratio α=√(M/N), and the average connectivity of the noise. The spectral observables are expressed in terms of a system of Recursive Distributional Equations, which are efficiently solved via a Population Dynamics algorithm. They allow us to compute the average largest eigenvalue ⟨λ1⟩A, the average top eigenvector component density, and the average overlap between the top eigenvector of A and x. We identify a critical threshold θcrit--depending on the average connectivity of the noise--that marks a BBP-like phase transition: below this value, ⟨λ1⟩A is unaffected by the signal, and the overlap vanishes. Thus, the signal is not recoverable from the top eigenvector of A. For θ>θcrit, the signal-related outlier eigenvalue becomes ⟨λ1⟩A and the overlap is nonzero, allowing for recovery of the signal. The results are in excellent agreement with numerical diagonalisation. We show that in the dense limit, the recovery threshold and eigen-statistics converge to the results predicted by the classical BBP transition for additive rank-one deformations of dense Wishart matrices.
Related
- Scaled-Gradient Hard Thresholding Algorithms for Sparse Signal Recovery
- Sparse Signal Recovery Using Log-Sum Regularization and Adaptive Smoothing
- The Noise Is the Signal: Correlated Sampling Error Is Rank-Informative for Proxy Metric Selection
- Low Rank for Rank: Uncertainty-Aware Task-Specific LLM Ranking under Sparse Pairwise Comparisons
- Effective DoF Collapse in Sparse Near-Field Beamforming: A Block-Rank-1 Mechanism
BibTeX
@misc{forer2026rankone,
title = {{Rank-One Signal Recovery in Sparse Wishart Noise}},
author = {Preben Forer and Urte Adomaityte and Pierpaolo Vivo},
howpublished = {arXiv},
year = {2026},
month = sep,
doi = {10.48550/arxiv.2609.28163},
eprint = {2609.28163},
archivePrefix = {arXiv},
url = {https://arxiv.org/abs/2609.28163},
}
RIS
TY - UNPB
TI - Rank-One Signal Recovery in Sparse Wishart Noise
AU - Preben Forer
AU - Urte Adomaityte
AU - Pierpaolo Vivo
PY - 2026
DA - 2026/09/23
PB - arXiv
DO - 10.48550/arxiv.2609.28163
UR - https://arxiv.org/abs/2609.28163
AB - We study the high-dimensional recovery of a signal vector x in the presence of sparse Wishart-like noise. We define an N × N matrix A = J+(θ/N)xx^⊤, where xx^⊤ is the rank-one deformation of the random noise matrix J. We consider a Wishart-like matrix J=X^⊤ X, where X is a sparse M × N random matrix with entries X_ij = c_ijW_ij, with c_ij regulating the density of non-zero elements, and W_ij the bond weights. Using the replica method, we compute analytically the top eigenpair statistics of A, and their dependence on the signal strength θ, the rectangularity ratio α=√(M/N), and the average connectivity of the noise. The spectral observables are expressed in terms of a system of Recursive Distributional Equations, which are efficiently solved via a Population Dynamics algorithm. They allow us to compute the average largest eigenvalue ⟨λ_1⟩_A, the average top eigenvector component density, and the average overlap between the top eigenvector of A and x. We identify a critical threshold θ_crit--depending on the average connectivity of the noise--that marks a BBP-like phase transition: below this value, ⟨λ_1⟩_A is unaffected by the signal, and the overlap vanishes. Thus, the signal is not recoverable from the top eigenvector of A. For θ>θ_crit, the signal-related outlier eigenvalue becomes ⟨λ_1⟩_A and the overlap is nonzero, allowing for recovery of the signal. The results are in excellent agreement with numerical diagonalisation. We show that in the dense limit, the recovery threshold and eigen-statistics converge to the results predicted by the classical BBP transition for additive rank-one deformations of dense Wishart matrices.
ER -