宽带卫星移动通信系统无线链路传输技术研究

发布时间：2018-11-05 16:13

【摘要】：卫星移动通信系统具有组网灵活、覆盖面积广以及不受自然灾害影响等特点,但同时卫星系统功率受限,信道传输延时大,衰落模型特殊,整体信道环境与陆地移动通信系统物理特性差异较大。利用MIMO-OFDM技术,构建适应卫星通信场景的宽带卫星移动通信系统,实现全球无缝互联通信网具有十分重大的意义。本文研究宽带卫星移动通信系统无线链路传输技术。首先,研究了卫星移动通信系统的信道传输特性及信道模型。针对卫星移动通信系统传输的应用场景,讨论了 Ka频段信道的天气影响及多普勒频偏,并给出了相应的信道模型。随后,研究了宽带卫星无线传输系统中全向信号传输方案的设计及实现。该传输方案中,天线的实际发送信号由低维空时编码信号经过全向预编码生成。结合阵列天线的信道特征,推导了全向传输矩阵需要满足的三个条件并分别基于Zadoff-Chu序列和Golay序列给出了全向预编码的设计实例。在此基础上,以下行同步信道为例,研究了全向传输预编码在公共信道传输中的应用,最后基于NI的USRP平台给出了硬件实现方案。接着,研究了基于空时频变换域滤波的宽带卫星移动通信下行信道估计方法。在卫星移动通信系统中,利用阵列天线的空间信道特征,结合远场散射场景中波束域稀疏性的特点,提出了基于空域滤波的信道估计方法,其思想是将空间域接收信号进行DFT变换到波束域,利用信道能量集中在有限波束上的特点,将其余波束能量作为噪声进行滤波,可以有效提升信道估计的性能。在传统时频二维变换域滤波方法的基础上,提出了基于空时频三维变换域滤波的信道估计方法。最后结合DOA估计,提出了基于DOA补偿的空域滤波信道估计方法。仿真结果表明这些方法可以获得更为精准的信道估计。最后,研究了宽带卫星移动通信上行链路中的低复杂度迭代接收机。在传统MMSE迭代接收机中,针对迭代过程中LMMSE均衡矩阵求逆计算复杂度过高的问题,提出了两种快速计算方法,分别为QR分解法和Sherman-Morrison公式化简法。随后给出了两种快速算法与矩阵直接求逆算法的复杂度和数值计算精度的比较,仿真结果表明这两种算法在不影响误码率的前提下,降低了迭代算法的计算复杂度。最后详细讨论了三种常用的QR分解算法,分别为Householder变换、Givens旋转以及Gram-Schmidt正交化算法,并分别从计算复杂度、数值稳定性、计算精度和硬件实现的难易程度几个方面对上述算法进行了分析。
[Abstract]:The satellite mobile communication system has the characteristics of flexible networking, wide coverage and not affected by natural disasters, but at the same time, the power of the satellite system is limited, the channel transmission delay is large, and the fading model is special. The physical characteristics of the whole channel environment and the terrestrial mobile communication system are quite different. It is of great significance to construct a wideband satellite mobile communication system based on MIMO-OFDM technology and to realize a global seamless interconnected communication network. In this paper, the wireless link transmission technology of wideband satellite mobile communication system is studied. Firstly, the channel transmission characteristics and channel model of satellite mobile communication system are studied. In this paper, the influence of weather and Doppler frequency offset on Ka channel is discussed, and the corresponding channel model is given for the application scenario of satellite mobile communication system. Then, the design and implementation of omnidirectional signal transmission scheme in wideband satellite wireless transmission system are studied. In this transmission scheme, the actual transmit signal of antenna is generated by omnidirectional precoding of low dimensional space-time coded signal. Based on the channel characteristics of array antenna, three conditions of omnidirectional transmission matrix are derived, and the design examples of omnidirectional precoding are given based on Zadoff-Chu sequence and Golay sequence respectively. On this basis, the application of omnidirectional transmission precoding in common channel transmission is studied by taking the following line synchronization channel as an example. Finally, the hardware implementation scheme based on the USRP platform of NI is given. Then, the downlink channel estimation method of wideband satellite mobile communication based on space-time-frequency transform domain filtering is studied. In the satellite mobile communication system, a channel estimation method based on spatial filtering is proposed based on the spatial channel characteristics of array antennas and the characteristics of beamspace sparsity in far-field scattering scenarios. The idea is to transform the received signal in spatial domain into beamspace by DFT. The channel energy is concentrated on the limited beam and the other beam energy is filtered as noise, which can effectively improve the performance of channel estimation. Based on the traditional time-frequency two-dimensional transform domain filtering method, a channel estimation method based on space-time-frequency three-dimensional transform domain filtering is proposed. Finally, combining with DOA estimation, a spatial filtering channel estimation method based on DOA compensation is proposed. Simulation results show that these methods can obtain more accurate channel estimation. Finally, the low complexity iterative receiver in the uplink of wideband satellite mobile communication is studied. In the traditional MMSE iterative receiver, in order to solve the problem of high computational complexity of LMMSE equalization matrix inversion, two fast computing methods, QR decomposition method and Sherman-Morrison formulaic simplified method, are proposed. Then the complexity and numerical accuracy of the two fast algorithms are compared with those of the direct matrix inversion algorithm. The simulation results show that the two algorithms reduce the computational complexity of the iterative algorithm without affecting the bit error rate (BER). Finally, three common QR decomposition algorithms are discussed in detail, which are Householder transform, Givens rotation and Gram-Schmidt orthogonalization. The calculation accuracy and the degree of difficulty in hardware implementation are analyzed.
【学位授予单位】：东南大学
【学位级别】：硕士
【学位授予年份】：2017
【分类号】：TN927.23

【相似文献】