摘要
认知无线电系统中,在频谱混叠或者缺少部分先验信息的条件下,能否将源信号分离出来决定了其通信质量的好坏。文中对一维均匀线性阵列结构下的混合信号模型,利用均匀线性阵列构成的混合矩阵具有较为特殊的计算特征,对该阵列输出进行矩阵分解和应用ESPRIT算法则获得对应的分离矩阵,实现混合信号的分离。相对经典Fast ICA算法分离计算复杂度有明显的降低。仿真实验说明,基于一维均匀线性阵列的信号分离具有比较好的效果。
In cognitive radio system,the mixed signal which has spectrum aliasing or lack of source signal prior information can be separated from the source signal also determines the communication quality. The paper bases on the one-dimensional mixed signal model of linear array structure,hybrid matrix using a uniform linear array has special characteristics calculation,the separationof mixed signals is realized by matrix decomposition and ESPRIT algorithm. The relative separation of classical Fast ICA algorithm has a lower computational complexity. The simulation results show that the signal separation based on one-dimensional uniform linear array has a better effect.
出处
《信息技术》
2018年第2期5-10,共6页
Information Technology
基金
华中师范大学中央高校基本科研业务费用项目资助(CCUN14A02005)
关键词
认知无线电
波达方向估计
信号分离
均匀线性阵列
cognitive radio
direction-of-arrival estimation
signal separation
uniform linear array
