稳健的对角加载广义内积算法

Robust generalized inner products algorithm using diagonal loading

针对空时自适应处理(STAP)在非均匀条件下的协方差矩阵估计问题,提出了对角加载广义内积(DL-GIP)算法.将对角加载用于改善广义内积(GIP)检测算法的检测性能.对于DL-GIP算法的关键是选择加载电平,给出了具体的选择方法.从理论上详细分析了对角加载对GIP性能的影响,即合理的对角加载可以提高检测概率和输出信噪比.最后的仿真分析验证了理论分析的正确性,即DL-GIP具有较高的检测概率和稳健的多个离群点检测能力,而且检测性能优于其他(广义内积和投影设计)非均匀检测(NHD)方法.

For the problem of the covariance matrix estimation in the nonhomogeneous condition for space-time adaptive processing （STAP）, the diagonal loading generalized inner products （DL-GIP） technique is proposed. By analyzing the GIP algorithm deeply, the diagonal loading is used to improve the GIP detection performance, and for DL-GIP, the key is to select the loading level, and the selecting method is advanced particularly. It is important that the effect of the diagonal loading upon the GIP performance be analyzed particularly in theory, namely the rational diagonal loading can improve the detection probability and the output signal-to-noise ratio. The last simulation attests its correctness and effectiveness, namely DL-GIP provides a higher probability of detection and robustness multiple outliers detection technique, and its performance is superior to other NHD methods this paper has mentioned.