基于AWD和EBD的认知雷达序贯信号检测

Sequential signal detecting for cognitive radar based on AWD and EBD

研究了认知雷达在扩展目标假设下,目标冲激响应未知时的信号检测问题。根据认知雷达闭环结构的特点,基于最大似然(ML)准则完成了对目标冲激响应的序贯估计,并结合自适应波形设计(AWD)和检测前估计(EBD)方法,提出了一种迭代更新目标冲激响应估计和发射波形的序贯检测算法。由理论分析和仿真结果可知,该算法对目标冲激响应的估计精度以及在近似恒虚警检测(CFAR)准则下的实际检测概率,均随着认知雷达循环次数的增加而提高,且估计的检测概率随着目标冲激响应估计精度的提高逐渐逼近实际检测概率,从而有效提高了决策的可靠性。

The problem of cognitive radar detecting for extended target with unkonwn target impulse response is studied.The sequential algorithm for estimating target impulse response based on maximum likelihood criterion is proposed according to the close-loop structure of cognitive radar,and considering the adaptive waveform design（AWD） and the estimate-before-detect（EBD） method,a new sequential detecting algorithm which updates the estimate of target impluse response and detecting waveform iteratively is proposed.The theoretical analysis and simulation results show that： if the proposed algorithm is used to detect target,the performance of estimating and detecting will be improved when the number of loop increases.And the estimated detecting probability will approach the true detecting probability when the estimating accuracy of target impulse response increases,which would improve the reliability of decision.