In urban environments, the challenges brought about by multipath propagation are significant for traditional adaptive radar detection problems. To tackle this issue, a generalized multipath signal model combined with environmental geometric prior information (EGPI) is developed. This model can be utilized regardless of whether the multipath signals are fully, partially, or partially-resolvable in the range domain. During the process of designing a detector using the criteria of the one-step generalized likelihood ratio test (GLRT), an attempt to find a closed-form expression for the maximum likelihood estimator of the unknown parameters was unsuccessful. However, by utilizing an adaptive configuration for the initial value, a cyclic optimization algorithm was able to obtain the estimated value within just a few iterations, demonstrating a rapid convergence rate. Furthermore, a detector based on the two-step GLRT criteria is proposed. Compared to traditional detectors, the newly developed detectors make more efficient use of the energy in multipath signals, resulting in improved detection performance. It has been demonstrated that when the whitened angles between subspaces are set to two specific values, the GLRT detectors exhibit the Constant False Alarm Probability (CFAR) property. Simulation results show that the false alarm probability() at these two specific angles provides upper and lower bounds of .

中文翻译：

---

知识辅助的广义多径自适应检测器

在城市环境中，多径传播带来的挑战对于传统的自适应雷达检测问题来说是显着的。为了解决这个问题，开发了结合环境几何先验信息（EGPI）的广义多径信号模型。无论多径信号在距离域中是完全、部分还是部分可解析，都可以利用该模型。在使用一步广义似然比检验 (GLRT) 的标准设计检测器的过程中，试图找到未知参数的最大似然估计量的封闭式表达式，但没有成功。然而，通过利用初始值的自适应配置，循环优化算法能够在几次迭代内获得估计值，表现出快速的收敛速度。此外，还提出了一种基于两步GLRT准则的检测器。与传统探测器相比，新开发的探测器可以更有效地利用多径信号中的能量，从而提高了探测性能。已经证明，当子空间之间的白化角度设置为两个特定值时，GLRT 探测器表现出恒定虚警概率 (CFAR) 属性。仿真结果表明，这两个特定角度的误报概率（）提供了 的上限和下限。