一种联合信道估计与多用户检测的新型算法
    点此下载全文
引用本文:陈发堂,石贝贝,杨恒.一种联合信道估计与多用户检测的新型算法[J].南京邮电大学学报:自然科学版,2020,40(3):15~21
摘要点击次数: 104
全文下载次数: 53
作者单位
陈发堂 重庆邮电大学 通信与信息工程学院重庆400065 
石贝贝 重庆邮电大学 通信与信息工程学院重庆400065 
杨恒 重庆邮电大学 通信与信息工程学院重庆400065 
基金项目:国家自然科学基金(61701063)资助项目
中文摘要:免调度非正交多址接入技术可以减少信令开销、以过载的方式容纳更多用户而受到人们的广泛关注。针对该技术需解决的信道估计、活跃用户检测和信号检测3个主要应用难题,文中提出一种联合信道估计的稀疏多用户检测方法同时解决上述3个问题。该方法通过联合考虑包含导频信号和用户数据信号的帧结构,重组用户信号为块稀疏单测量向量的压缩感知模型,将问题转化为块稀疏信号恢复。所提算法通过验证误差的方式自适应地结束迭代,相比于根据不同的信噪比而设置对应迭代停止阈值的块稀疏子空间追踪算法灵活性更好。仿真结果表明,所提算法能够高可靠地完成联合信道估计和多用户检测,相比于独立的信道估计和多用户检测误码率性能得到了很大的提升,且算法灵活性好,更适用于实际的通信系统。
中文关键词:非正交多址接入  压缩感知  联合信道估计和多用户检测
 
Novel joint channel estimation and multi user detection algorithm
Abstract:A grant free non orthogonal multiple access technology can reduce the signaling overhead and accommodate more users in an overload way. Aiming at the three main application problems of channel estimation, active user detection and signal detection , a sparse multi user detection method for the joint channel estimation is proposed to solve the above three problems. Combined the frame structure containing pilot signal with user data signal, a compressed sensing model for block sparse single measurement vectors is established to transform the problem into block sparse signal recovery. The algorithm adaptively ends iterations by verifying errors, thus it is more flexible than the block sparsity adaptive subspace pursuit algorithm that sets the corresponding iteration stop threshold according to the different signal to noise ratios. Simulation results show that the algorithm can complete the joint channel estimation and the multi user detection with high reliability. Compared with independent channel estimation and multi user detection, the performance of the symbol error rate is improved. The algorithm is flexible and more suitable for the actual communication system.
keywords:non orthogonal multiple access  compressive sensing  joint channel estimation and multi user detection
查看全文  查看/发表评论  下载PDF阅读器

你是第3044537访问者
版权所有《南京邮电大学学报(自然科学版)》编辑部
Tel:86-25-85866913 E-mail:xb@njupt.edu.cn
技术支持:本系统由北京勤云科技发展有限公司设计