A Semianalytical Framework for Noncoherent Detection Performance Evaluation of Targets with Compound Fluctuations

Evaluating radar detection performance for complex, nonstationary targets is challenging. To address this problem, we propose a semianalytical detection performance evaluation framework for noncoherent radars that leverages the efficiency of analytical methods and the adaptability of statistical simulations. First, we formulate the framework by adopting a compound fluctuation model to decompose the detection probability into tractable conditional terms. Then, within this framework, we develop an adaptive semianalytical (ASA) method, which uses adaptive importance sampling to efficiently approximate the conditional probability summation over the joint parameter space without exhaustively enumerating all parameter combinations. Next, we derive a theoretical upper bound on the estimation error of ASA, providing guarantees on its reliability. Finally, we validate the proposed framework and ASA through numerical experiments on representative scenarios and a real-data example, showing that ASA achieves lower root mean square error than conventional methods. The proposed framework and ASA thus provide an accurate and efficient tool for noncoherent detection performance evaluation in advanced radar system design.