一种新的稳健匹配子空间检测算法研究

摘要

考虑了一种特殊情况下的白高斯噪声下的稳健匹配子空间检测算法，即假定在备择假设下噪声功率已知，而零假设下的噪声功率未知。首先推导了基于广义似然比的检验统计量，该检测统计量的分布可以近似为F分布与χ2分布2个独立分量的线性组合形式，给出了适合数值计算的检测性能计算公式，最后利用计算机仿真讨论了该算法的检测性能，并验证了理论分析的有效性。%In this paper, we consider a robust matched subspace detection algorithm under a particular condition that the noise power is supposed to be known under alternative hypothesis, and to be unknown under null hypothesis.The test statistic is derived based on generalized likelihood ratio test, and can be approximately expressed as the linear combination of two independ-ent components, which are F distribution and χ2 distribution respectively.The formula for numerical computation is given, and the computer simulation is used to analyzed the detection performance of the algorithm, and to validate the theory results.