摘要
针对密度聚类方法普遍存在参数调整复杂性、无法有效识别子簇的整体几何特性问题,设计了一种基于持续同调技术的识别方法。该方法首先提取星座图的一阶拓扑特征,根据是否具有显著的一阶拓扑特征区分多进制数字相移键控(MPSK)和多进制正交幅度调制(MQAM);然后综合零阶、一阶拓扑特征,计算出持续时间足够长的单连通分支数进行两类信号的类内识别。仿真实验表明,所提方案在加性白高斯噪声(AWGN,0~30 dB)环境下,MQAM和MPSK二分类的正确识别率达到了97.40%,类内识别准确率高于传统的聚类方案。将所提算法在基于通用软件无线电设备(USRP)进行收发的信号上进行测试,验证了该方法的有效性。
To address the two major problems commonly associated with density clustering methods:the complexity of parameter tuning and the inability to effectively recognize the overall geometric properties of subclusters,a recognition method based on Persistent Homology(PH)is designed.The method first extracts the 1st-order topological features of the constellation diagram and distinguishes between Multiple Phase Shift Keying(MPSK)and Multiple Quadrature Amplitude Modulation(MQAM)based on whether they have significant 1st-order topological features;then it synthesizes the 0th-order and 1st-order topological features,and calculates the number of single-connecting branches with a sufficiently long duration to perform the intra-class identification of the two types of signals.Simulation experiments show that the proposed scheme achieves a correct recognition rate of 97.40%for MQAM and MPSK binary classification in Adaptive White Gaussian Noise(AWGN)(0~30 dB)environment,and intra-class recognition accuracy is higher than that of traditional clustering schemes.Finally,the proposed algorithm is tested on signals sent and received based on Universal Software Radio Peripheral(USRP)to verify the effectiveness of the method in the paper.
作者
王溪
孟华
代彬
周淳
WANG Xi;MENG Hua;DAI Bin;ZHOU Chun(School of Information Science and Technology,Southwest Jiaotong University,Chengdu Sichuan 611756,China;School of Mathematics,Southwest Jiaotong University,Chengdu Sichuan 611756,China)
出处
《太赫兹科学与电子信息学报》
2025年第9期966-975,983,共11页
Journal of Terahertz Science and Electronic Information Technology
关键词
星座图
调制识别
持续同调
正交幅度调制/相移键控(QAM/PSK)
constellation diagram
modulation classification
Persistent Homology(PH)
Quadrature Amplitude Modulation/Phase Shift Keying(QAM/PSK)
