针对Alamouti空时块编码复用增益损失的问题,提出了两组Alamouti编码方案。在此基础上,为了改善系统的误码率(BER)性能和简化接收端复杂度,提出了将几何均值分解(GMD)算法和非线性预编码技术相结合的两组Alamouti传输方案。该方案的设...针对Alamouti空时块编码复用增益损失的问题,提出了两组Alamouti编码方案。在此基础上,为了改善系统的误码率(BER)性能和简化接收端复杂度,提出了将几何均值分解(GMD)算法和非线性预编码技术相结合的两组Alamouti传输方案。该方案的设计方法为:首先等效出两组Alamouti空时块编码系统的信道矩阵;进而,通过GMD算法对等效信道矩阵进行收发端联合设计;最后,在发射端应用脏纸(dirty paper coding,DPC)和Tomlinson-Harashima precoding(THP)非线性预编码技术,消除发送信号间的干扰,从而使系统获得更好的误码率性能。通过仿真结果对比发现,提出的系统可以显著地改善误码率性能。展开更多
考虑使用深度神经网络(DNN)开发一种决策导向(DD)信道估计(CE)算法,用于高度动态车辆环境中的MIMO空时块编码系统。Mehrtash Mehrabi等人2019年11月在《IEEE Journal on Selected Areas in Communications》发表文章,提出利用DNN进行空...考虑使用深度神经网络(DNN)开发一种决策导向(DD)信道估计(CE)算法,用于高度动态车辆环境中的MIMO空时块编码系统。Mehrtash Mehrabi等人2019年11月在《IEEE Journal on Selected Areas in Communications》发表文章,提出利用DNN进行空时块码(STBC)k步信道预测,并证明了基于深度学习(DL)的DD-CE算法不再需要估计快速时变准静态信道中随包不断变化的多普勒调频率。在这种车载信道中,估计多普勒频率的挑战性很大,需要大量的导频和前导码,从而导致功率和频谱效率较低。作者训练了两个深度神经网络,它们学习多普勒调频率范围很大的情况下MIMO衰落信道的实部和虚部。训练证明,通过以上深度神经网络,仅需多普勒调频率范围的先验知识,而无需知道其确切值,就可实现DD-CE。展开更多
Space-Time Block (STB) code has been an effective transmit diversity technique for combating fading due to its orthogonal design, simple decoding and high diversity gains. In this paper, a unit-rate complex orthogonal...Space-Time Block (STB) code has been an effective transmit diversity technique for combating fading due to its orthogonal design, simple decoding and high diversity gains. In this paper, a unit-rate complex orthogonal STB code for multiple antennas in Time Division Duplex (TDD) mode is proposed. Meanwhile, Turbo Coding (TC) is employed to improve the performance of proposed STB code further by utilizing its good ability to combat the burst error of fading channel. Compared with full-diversity multiple antennas STB codes, the proposed code can implement unit rate and partial diversity; and it has much smaller computational complexity under the same system throughput. Moreover, the application of TC can effectively make up for the performance loss due to partial diversity. Simulation results show that on the condition of same system throughput and concatenation of TC, the proposed code has lower Bit Error Rate (BER) than those full-diversity codes.展开更多
文摘针对Alamouti空时块编码复用增益损失的问题,提出了两组Alamouti编码方案。在此基础上,为了改善系统的误码率(BER)性能和简化接收端复杂度,提出了将几何均值分解(GMD)算法和非线性预编码技术相结合的两组Alamouti传输方案。该方案的设计方法为:首先等效出两组Alamouti空时块编码系统的信道矩阵;进而,通过GMD算法对等效信道矩阵进行收发端联合设计;最后,在发射端应用脏纸(dirty paper coding,DPC)和Tomlinson-Harashima precoding(THP)非线性预编码技术,消除发送信号间的干扰,从而使系统获得更好的误码率性能。通过仿真结果对比发现,提出的系统可以显著地改善误码率性能。
文摘考虑使用深度神经网络(DNN)开发一种决策导向(DD)信道估计(CE)算法,用于高度动态车辆环境中的MIMO空时块编码系统。Mehrtash Mehrabi等人2019年11月在《IEEE Journal on Selected Areas in Communications》发表文章,提出利用DNN进行空时块码(STBC)k步信道预测,并证明了基于深度学习(DL)的DD-CE算法不再需要估计快速时变准静态信道中随包不断变化的多普勒调频率。在这种车载信道中,估计多普勒频率的挑战性很大,需要大量的导频和前导码,从而导致功率和频谱效率较低。作者训练了两个深度神经网络,它们学习多普勒调频率范围很大的情况下MIMO衰落信道的实部和虚部。训练证明,通过以上深度神经网络,仅需多普勒调频率范围的先验知识,而无需知道其确切值,就可实现DD-CE。
基金Supported by Chinese 863 project (No.2001 AA 123042).
文摘Space-Time Block (STB) code has been an effective transmit diversity technique for combating fading due to its orthogonal design, simple decoding and high diversity gains. In this paper, a unit-rate complex orthogonal STB code for multiple antennas in Time Division Duplex (TDD) mode is proposed. Meanwhile, Turbo Coding (TC) is employed to improve the performance of proposed STB code further by utilizing its good ability to combat the burst error of fading channel. Compared with full-diversity multiple antennas STB codes, the proposed code can implement unit rate and partial diversity; and it has much smaller computational complexity under the same system throughput. Moreover, the application of TC can effectively make up for the performance loss due to partial diversity. Simulation results show that on the condition of same system throughput and concatenation of TC, the proposed code has lower Bit Error Rate (BER) than those full-diversity codes.
