Enabling Real-Time Imaging and Onboard RFI Localization for Three-Level Quantized Microwave Interferometric Radiometers

Real-time imaging processing for microwave interferometric radiometer (MIR) has great potential in various application fields, such as onboard data processing, onboard information fusion, and alternative visual applications. The primary challenge lies in the computational complexity of the entire processing chain, including both visibility function preprocessing and brightness temperature (TB) reconstruction. In this study, the real-time estimation of the normalized threshold level is identified as the key step for enabling real-time imaging of three-level quantized MIR systems. Three algorithms—Acklam’s algorithm (AKA), polynomial fitting algorithm (PFA), and Taylor expansion algorithm (TEA)—are proposed and evaluated. The PFA provides a favorable balance between estimation accuracy and computational efficiency. Leveraging the proposed algorithms, this paper further establishes an onboard real-time processing framework for three-level quantization MIRs, enabling real-time TB imaging and radio frequency interference (RFI) localization. A real-time imaging experiment was carried out with a 15-element, 50 GHz one-dimensional MIR system, which demonstrates real-time imaging of fast-moving vehicles on the expressway with greatly reduced computational latency (an imaging time of 570.9 μs for 159 baselines). A further flight experiment employing an L-band system verifies the feasibility of onboard RFI localization, and the proposed real-time RFI localization method shows an average angular deviation of 0.32° with respect to an offline MUSIC estimator, corresponding to 2.1% of the nominal spatial resolution.