摘要
波形设计直接关系到MIMO雷达目标检测和参数估计的性能,通过设计发射互相关矩阵来分配空间发射功率是一类MIMO雷达波形设计方法。本文在发射方向图综合的二次规划(Quadratic Programming,QP)模型基础上,提出了一种基于DFP拟牛顿法的MIMO雷达方向图综合设计方法,该方法用可变的一维搜索步长代替了最陡下降法的固定步长,并且不需计算Hesse逆矩阵。仿真结果表明,该方法复杂度低,较最陡下降法收敛速度快,并具有更好的数值稳定性,可以满足MIMO雷达发射方向图设计的要求。
Waveform design affects the performance of target detection and parameters estimation for MIMO radar. The transmit power can be assigned by the design of the transmit cross-correlation matrix. On the basis of the QP model of transmit pattern synthesis, a algorithm for MIMO radar pattern design based on DFP Newton-like method is presened in this paper. The algorithm replaces the fixed searching length of steepest descent method with changing one-dimensional searching length without the calculation of the inverse of Hesse matrix. The simulation result indicates that the new algorithm is preferable for its low complexity, better convergence and stability compared with the steepest descent method and consequently it has the advantage in meeting the demands of the design of transmit pattern for MI- MO radar.
出处
《电波科学学报》
EI
CSCD
北大核心
2008年第6期1188-1193,共6页
Chinese Journal of Radio Science
基金
国防科技重点实验室基金项目(ZZ0303072208)
关键词
MIMO雷达
波形设计
二次规划
拟牛顿法
MIMO radar
waveform design
quadratic programming
Newton-like method
