异构网络干扰信道估计研究

发布时间：2019-06-25 09:30

【摘要】：城市化发展带来的整体环境的改变,使得无线信号的传播信道愈加恶劣,与此同时,室内数据、语音等业务需求不断增大,这使得高质量的室内覆盖变得十分迫切。家庭基站作为一种室内覆盖技术,有效提升了整体网络的容量,不仅发射功率低,而且部署方便。家庭基站和宏基站构成了双层异构网络,二者使用相同的频谱资源,相较于传统网络,异构网络的结构复杂,干扰也更加严重。异构网络中得到干扰信道信息后,可以进行干扰消除。协作网络中利用导频信息通过LS、MMSE算法进行干扰信道估计,但在非协作网络中,当期望信道的数据信息和干扰信道的导频信息重叠时,利用LS、MMSE算法得到的干扰信道估计性能较差。本文基于无线宽带信道的稀疏性质,在非协作异构网络的干扰信道估计中,提出一种基于压缩感知的迭代方法,通过迭代估计得到干扰信道信息和期望数据信息。文中建立了双层异构网络的下行链路系统仿真模型,对比采用压缩感知和采用LS的干扰信道估计性能。仿真结果表明,利用压缩感知技术可准确估计异构网络中的干扰信道,路径数量发生变化时,估计的性能受到部分影响,当SNR高于19dB时,基于压缩感知的干扰信道估计性能明显优于LS算法。本文在异构网络中基于压缩感知对干扰信道估计性能进行优化,提出了两种优化方法:(1)通过构造确定性稀疏二值观测矩阵,获得相应的感知矩阵,基于OMP优化恢复算法;(2)给定稀疏基,通过求解最小Frobenius范数,使观测矩阵具有较小的互相关值,优化观测矩阵。仿真结果表明,利用优化算法能够有效降低干扰信道估计的均方误差0.5-2dB。
[Abstract]:With the change of the overall environment brought by the development of urbanization, the propagation channel of wireless signal is becoming worse and worse. at the same time, the demand for indoor data, voice and other services is increasing, which makes high-quality indoor coverage very urgent. As an indoor coverage technology, home base station effectively improves the capacity of the whole network, which not only has low transmission power, but also is convenient to deploy. Home base station and macro base station constitute double-layer heterogeneous network, which use the same spectrum resources. Compared with the traditional network, the structure of heterogeneous network is more complex and the interference is more serious. Interference cancellation can be carried out after interference channel information is obtained in heterogeneous networks. In cooperative networks, pilot information is used to estimate interference channels by LS,MMSE algorithm, but in non-cooperative networks, when the data information of expected channels and pilot information of interference channels overlap, the performance of interference channel estimation obtained by LS,MMSE algorithm is poor. In this paper, based on the sparse nature of wireless broadband channel, an iterative method based on compressed sensing is proposed in the interference channel estimation of non-cooperative heterogeneous networks, and the interference channel information and expected data information are obtained by iterative estimation. In this paper, a downlink system simulation model for double-layer heterogeneous networks is established, and the interference channel estimation performance using compressed sensing and LS is compared. The simulation results show that the interference channel in heterogeneous networks can be accurately estimated by using compressed sensing technology. When the number of paths changes, the performance of the estimation is partly affected. When the SNR is higher than 19dB, the performance of interference channel estimation based on compression sensing is obviously better than that of LS algorithm. In this paper, the performance of interference channel estimation is optimized based on compressed sensing in heterogeneous networks, and two optimization methods are proposed: (1) the corresponding perception matrix is obtained by constructing deterministic sparse binary observation matrix, and the corresponding perceptual matrix is optimized based on OMP; (2) given sparse basis, the observation matrix has a small cross-correlation value and optimizes the observation matrix by solving the minimum Frobenius norm. The simulation results show that the mean square error of interference channel estimation can be effectively reduced by using the optimization algorithm.
【学位授予单位】：南京邮电大学
【学位级别】：硕士
【学位授予年份】：2017
【分类号】：TN929.5

【参考文献】