摘要
针对毫米波大规模多入多出系统在射频链路数量有限时,波束域信道估计是一个有挑战性的问题,提出一种基于优化BM3D的信道估计方案。利用基于三维透镜的多入多出系统信道矩阵可被视为二维自然图像的结构特性,将图像重构技术融入信道估计。BM3D是目前最精确的图像去噪算法之一,通过块匹配实现分组,利用三维变换域的收缩完成协同滤波。考虑信道的稀疏特性和路径的聚类特性,对BM3D算法进一步优化以提高性能。仿真结果表明,提出的优化BM3D方案在所有考虑的信噪比区域均能取得令人满意的信道估计精度。
Focused on the issue in which beamspace channel estimation is challenging for mmWave massive MIMO system when the number of RF chains is limited,a channel estimation scheme based on optimized BM3D was proposed.The MIMO channel matrix based on 3D lens-based can be regarded as a 2D natural image,and the image reconstruction technology was integrated into the channel estimation.BM3D was considered as one of the most accurate algorithms for image denoising,while the grouping was realized by block-matching and collaborative filtering was accomplished by shrinkage in a 3D transform domain.Utilizing the sparsity feature of the channel and the clustering feature of the paths,the BM3D algorithm was optimized in order to improve the performance.Simulations are provided to show that proposed optimized BM3D scheme can achieve satisfactory accuracy in all considered SNR regions.
作者
邱佳锋
QIU Jiafeng(School of Communication Engineering,Hangzhou Dianzi University,Hangzhou 310018,China)
出处
《电信科学》
2020年第9期44-50,共7页
Telecommunications Science
基金
浙江省自然科学基金资助项目(No.LY12F01008)。
