论文部分内容阅读
对信号脉冲重复频率(PRF)的调制样式进行自动识别是信息处理领域的关键技术之一。为解决PRF调制识别算法识别样式有限、准确率低等问题,以雷达信号为例,提出了一种新的识别算法。对接收到的信号序列分别提取比值特征、比重特征、频率特征和形状特征4个参数,基于径向基概率神经网络,实现对多种信号PRF调制样式的自动识别。仿真结果表明:在脉冲丢失率和脉冲虚警率均为10%的情况下,利用该方法可以得到最低为90%的识别准确率,能够满足当前的电磁环境,具有一定的工程参考价值。
Automatic modulation of the signal pulse repetition frequency (PRF) modulation pattern is one of the key technologies in the field of information processing. In order to solve the problem of limited recognition pattern and low accuracy of PRF modulation recognition algorithm, a new recognition algorithm is proposed by taking radar signal as an example. Four parameters of ratio, gravity, frequency and shape were extracted from the received signal sequence respectively. Based on radial basis probabilistic neural network, the PRF modulation patterns of various signals were automatically identified. The simulation results show that under the conditions of both the pulse loss rate and the impulsive false alarm rate are 10%, the method can get the recognition accuracy of 90% at the minimum, which can meet the current electromagnetic environment and has certain engineering reference value.