Rapid Excitation Retrieval for Large-Scaled Antenna Arrays Using Nonconvex Optimization and Phaseless Measurements

We present a novel method for excitation retrieval in large phased arrays using phaseless single-probe measurements. Non-convex optimization with both phaseless and single-probe measurements is integrated into a unified framework for practical excitation reconstruction. Our method employs a fixed measurement probe and does not require prior knowledge of the excitation or aperture field. A truncated amplitude flow (TAF) algorithm is developed, featuring two key enhancements: an orthogonality-promoting initialization and a gradient truncation strategy to avoid local minima, both of which significantly improve retrieval accuracy and robustness. Numerical simulations on various array configurations demonstrate that the proposed method offers superior speed, accuracy, and noise resilience compared to existing convex and non-convex approaches. The practical effectiveness of the algorithm is further validated through experimental measurements on a 64-element aperiodic phased array using receiving probes with different polarization states.