论文部分内容阅读
在计算雷达和声纳系统的检测概率时,通常是假定给出的某个虚警概率所需的门限是已知的。这种假定往往并非是真实情况,因此,必须进行某些试验统计测量来估算门限。本文介绍了在一般结构中处理估算(自适应)门限的计算技术,该结构对解决许多检测问题都能适应。本文根据噪声或信号加噪声变化的特征函数和门限估算的特征函数来计算虚警和检测概率;同时为了说明这种方法,对窄带高斯目标信号,本文计算出接叠离散傅里叶变换(DFT)的检测性能。
When calculating the detection probabilities for radar and sonar systems, it is usually assumed that the threshold required for a given false alarm probability is known. This assumption is often not the real situation, therefore, some experimental statistical measurements must be made to estimate the threshold. This article describes computational techniques for handling the estimation (adaptive) threshold in a general structure that is adaptable to solve many detection problems. In this paper, the false alarm and the detection probability are calculated according to the noise-signal or signal-plus-noise eigenfunction and the threshold-based eigenfunction estimation. In order to illustrate this method, for the narrow-band Gaussian target signal, the overlapped discrete Fourier transform ) Detection performance.