低位宽量化对大规模多天线系统可达速率的影响研究

发布时间：2018-02-24 20:45

本文关键词： 非线性失真低位宽ADC 大规模多天线技术加性噪声模型线性接收机广义互信息　出处：《中国科学技术大学》2017年硕士论文　论文类型：学位论文

【摘要】：近些年移动互联网用户数和移动终端数量正快速增长,移动互联网用户对高清视频等高质量的服务的需求逐渐提高,这给现有的无线通信系统和技术带来了巨大的压力。为了缓解压力,下一代无线通信系统中使用的频率带宽和天线数量都会大幅增长,Massive MIMO、毫米波等技术大幅提高了能量的效率和频谱的效率,但这给通信系统的硬件设计引入了一些新的挑战。这些挑战中,有关于ADC的功率消耗就是一个急需解决的重要问题。低位宽的,特别是1-bit位宽的ADC彻底改变了通信的理论和应用。在本论文中,我们采用广义互信息来衡量信道输出端经过低位宽量化的Massive MIMO系统的性能。广义互信息的物理意义是系统在给定了输入的码元分布和解码方式下的速率:如果发送信号速率低于这个速率,全体码本统计意义上的平均解码错误概率随着码长趋于无穷变为0;相反地,如果发送信号速率高于这个速率,解码错误概率随着码长趋于无穷变为1。基于高斯码本和最近邻解码,我们得出了一个有关于通信系统收发端的失真的等效关系,这个等效关系验证了基于GMI的有效信噪比和基于Bussgang理论推导出的启发式的信噪比形式上相等。物理上,量化过程使用ADC来实现,但是ADC的非线性特性给分析带来了困难。加性量化噪声模型(AQNM)是一个广泛使用的近似描述量化过程的强分析性模型。所以我们采用加性量化噪声模型来近似分析系统速率,并与实际量化系统的可达速率进行对比。仿真结果显示,基于加性量化噪声模型的近似速率和量化系统的可达速率在单用户场景中差异较大,但是这个区别可以通过在加性量化噪声模型中考虑量化噪声矢量之间的相关性最终取得一致性。广义互信息的一般解析式涉及到接收机,量化输出矢量的协方差矩阵,量化输出矢量和发送信号的相关矢量。我们在量化系统中采用一种低复杂度的接收机,以降低系统开销。仿真结果表明这个低复杂度的次优接收机带来的性能增益在BER和系统速率角度均接近最优的接收机。
[Abstract]:In recent years, the number of mobile Internet users and the number of mobile terminals are growing rapidly, and the demand of mobile Internet users for high-quality services such as high-definition video is gradually increasing. This puts tremendous pressure on existing wireless communication systems and technologies. The frequency bandwidth and the number of antennas used in the next generation wireless communication systems will increase greatly, such as massive MIMOs, millimeter-wave and other technologies that greatly improve the energy efficiency and spectral efficiency. But this introduces some new challenges to the hardware design of the communication system. Among these challenges, the power consumption of ADC is an important problem that needs to be solved. In particular, 1-bit bit wide ADC has completely changed the theory and application of communication. We use generalized mutual information to measure the performance of Massive MIMO systems with low bandwidth quantization at the output end of the channel. The physical meaning of generalized mutual information is the rate of the system under the given input symbol distribution and decoding mode: if the transmitter is transmitted. The signal rate is lower than this rate. The average decoding error probability in the statistical sense of all codebooks tends to zero with the code length; conversely, if the transmission signal rate is higher than this rate, Based on Gao Si codebook and nearest neighbor decoding, we obtain an equivalent relation of distortion on the transceiver of communication system. The equivalent relation verifies that the effective SNR based on GMI is equivalent to the heuristic SNR derived from Bussgang theory. Physically, the quantization process is realized by ADC. But the nonlinear characteristics of ADC make it difficult to analyze. The additive quantization noise model is a strong analytical model which is widely used to describe the quantization process. So we use the additive quantization noise model to approximate the system rate. The simulation results show that the approximate rate based on additive quantization noise model and the reachable rate of quantization system differ greatly in single user scenario. But this difference can be achieved by considering the correlation between quantization noise vectors in the additive quantization noise model. The general analytic formula of generalized mutual information relates to the receiver, the covariance matrix of quantization output vector, Quantization output vectors and correlation vectors of transmitted signals. We use a low complexity receiver in the quantization system. The simulation results show that the performance gain of this low-complexity sub-optimal receiver is close to the optimal one in both the BER and system rate angles.
【学位授予单位】：中国科学技术大学
【学位级别】：硕士
【学位授予年份】：2017
【分类号】：TN92

【相似文献】