Standard Condition Number-Based Robust Signal Detection with Whitening under Uncertainty

TL;DR

This paper introduces a unified analytical framework for the Standard Condition Number detector, enhancing robustness and providing finite-sample performance analysis in colored noise environments with covariance uncertainty.

Contribution

It develops a comprehensive RMT-based analysis of the SCN detector's false-alarm and detection probabilities, including CFAR property under covariance mismatch.

Findings

SCN maintains CFAR property despite covariance mismatch

Analytical expressions for false-alarm and detection probabilities are derived

Simulation confirms improved robustness over traditional detectors

Abstract

Robust signal detection in colored noise with unknown covariance is essential in radar, cognitive radio, integrated sensing and communication (ISAC), and quantum sensing applications. This paper develops a unified analytical framework for the Standard Condition Number (SCN) detector, which employs the ratio of the largest to smallest eigenvalues of the whitened sample covariance matrix. The framework jointly covers both ideal conditions in which the training and sensing noise statistics are identical and disturbed conditions in which interference or jamming alters the sensing covariance. Despite the SCN's practical relevance, its finite-sample false-alarm and detection behavior has not been analytically characterized. Using random matrix theory (RMT), we derive general expressions for these probabilities, provide closed-form results for special cases, and show that the SCN preserves the Constant False Alarm Rate (CFAR) property under covariance mismatch. Analytical and simulation results confirm that the proposed unified framework delivers consistent detection performance and greater robustness than conventional eigenvalue- and LRT-based detectors.