基于k-means算法的直扩信号信息及伪码序列盲估计方法

In this paper, we study the blind estimation of information sequence and Pseudo-Noise sequence of Short Code-Direct Sequence Spread Spectrum (SC-DSSS) signals under non-cooperative conditions.In order to solve the problems of large computation and slow estimation speed of the existing matrix decomposition methods under the condition of long pseudo code period and large observation matrix, a new method of SC-DSSS signal information sequence and pseudo code sequence estimation based on K-means clustering was proposed.First, the sampled observation signals are segmented according to the width of information code, and then the segmented codes are clustered by K-means algorithm. In the case of unknown pseudo code sequence, the blind estimation of information sequence is realized.On this basis, the information chip is further segmented according to the width of the pseudo code, and the pseudo code is clustering to realize the blind estimation of pseudo code sequence.Simulation results show the feasibility and effectiveness of the proposed method.      

基于IQ图特征的小样本通信辐射源个体识别

在真实的战场环境中,我们很难采集到足够的带标签的敌方辐射源数据,因此,小样本学习变得越来越重要。通过不断地发展,CNN神经网络有着很强的处理图片分类的能力。在小样本条件下,为了充分利用发展最为成熟的CNN神经网络,本文提出了将一维IQ数据转化成二维的IQ图特征的方法,来进行针对小样本的分类任务。由于数据的IQ图具有重复性与个体的差异性,通过实验,这种方法在识别不同个体超短波电台上有着99.5%的正确率,对比双谱特征,IQ图特征具有更强的识别能力。这种特征变换方法简单,并且CNN网络处理图片分类的技术成熟,具有很强的实用性。      

基于特征值分解的直扩信号信息序列和伪码序列盲估计方法

本文研究了非协作扩频通信中短码直扩信号的盲解扩问题,针对现有方法在低信噪比条件下失步时间估计不准等问题,本文提出了一种改进的基于特征值分解的直扩信号信息和伪码序列盲估计方法。方法首先定义了第一类接收信号自相关矩阵,并根据其特征值分解的特点,利用矩阵特征值与失步点的关系完成信号失步点的估计,随后利用特征向量在伪码序列未知的情况下估计出信息序列。失步点估计完成以后,再利用第二类接收信号自相关矩阵的特征向量实现伪码序列的盲估计。实验结果表明方法具有估计精度高、稳定性好、抗噪性能强的优点。