基于AR模型的Rao检测算法

Rao detection algorithm based on AR modeling

将相关噪声看作为AR模型,在未知信号复幅度而已知目标多普勒频移情况下,利用Rao检验方法实现了雷达目标检测。该算法只在H0假设条件下对未知参数进行最大似然估计,并通过计算机仿真了该检测算法的检测性能。仿真结果表明该检测算法的检测性能与GLRT算法性能相当,而结构则更为简单,同时文中也给出了该检测算法的渐近性能。

A detection algorithm is derived to implement radar target detection through the Rao test with known Doppler frequency shift and unknown complex amplitude of the correlated noise, which is assumed as an autoregressive （AR） model. The unknown parameters are estimated by maximum likelihood estimation under hypothesis H0. It is shown through computer simulations that the algorithms has almost the same performance as that of the GLRT, while the structure is simpler. The asymptotic performance of the algorithms is also given.