The aim of this paper is to present a novel threshold setting algorithm by which both the false alarm rate and the probability of detection could be kept constant in a range while the SNR of received signals varies.What makes it different from the former adaptive threshold setting rules is that it is not based on the conventional CFAR criterion but on a novel constant equivalent signal to noise ratio(CESNR)criterion. Since only when the output SNR is constant can the probability of detection be kept stable along with a desired false alarm rate,the goal of the threshold detector is to maintain a constant output SNR by using the non-coherent combination.On one side,the adaptive threshold is set as a function of the estimatednoise power,thus guarantee the accumulated output that exceed the threshold would reach a required resulting signal to noise ratio.On the other side,the number of tested samples is adaptively adjusted,which means,the integration time is increased until the non-coherent correlation output exceed the threshold.Therefore,whenever a detection is declared,the resulting SNR is constant,so that the probability of detection and the false alarm rate could be kept constant no matter how much the input SNR varies.%针对GPS接收机中的信号捕获问题提出了一种新颖的判决门限的自适应设定方法,它在接收信号的信噪比大幅度变化的情况下也可同时保证捕获的虚警概率和检测概率在一定范围内稳定不变.与以往自适应门限设定方法不同的是,它不是基于传统的恒虚警概率判决准则,而是基于一种新颖的恒等效信噪比准则.因为只有当输出信噪比恒定时,才有可能在虚警概率恒定的同时也保证检测概率稳定,所以这个门限监测器的目的就是要通过非相干累加来实现累加后的等效信噪比的恒定.一方面,将自适应门限设置成实时估计的噪声能量的函数,从而保证超过门限的累加输出值一定能达到指定的等效信噪比.另一方面,检测样本数也是在自适应的调整,即累加时间会不断增加直到输出的非相干累加结果超过当时的门限值.所以,一旦判决为检测到信号时,这时的输出等效信噪比都是恒定的,因此检测概率和虚警概率都可保持稳定而无论当时的输入信噪比如何变化.
展开▼
