28GHz多接收器MIMO系统低复杂度混合波束形成

发布时间：2018-05-25 19:07

本文选题：毫米波 + 混合波束形成　；参考：《哈尔滨工业大学》2017年硕士论文

【摘要】：目前的无线服务已经饱和,几乎所有6GHz以下的频谱以及对高速无线通信的广泛增长的需求,特别是与沉浸式多媒体应用程序正在进入智能设备(例如智能手机,平板电脑,笔记本电脑等)。空间视频流量占移动流量的61%,预计未来几年将会快速增长。全高清视频的丰富性通过社交媒体分享,超高清和3D视频内容将会在未来发生,这促使研究人员深入研究了毫米波(mm波)频带作为经典乐队的替代品,并将广泛使用的带宽特权作为第五代无线通信系统及其以外即将到来的需求的有效解决方案。毫米波频段特别是28GHz,我们的研究兴趣仍然处于研究阶段。作为下一代无线通信的解决方案,毫米波频带中的一个显著的通道传播特性是通过频带的巨大传输信号路径损耗。以前的研究已经显示了毫米波的通道测量结果,其中与波段频谱差比传统的蜂窝系统频带更差的传播损耗特性。而且,它们也已经显示出基于波束形成技术的无线通信的可能性,其通过小波长的毫米波,对于视径(Lo S)和非视径(NLo S)发射机和接收机传播而变得更容易实现。已经证实,当我们从1.8GHz跳到28GHz再到60GHz频带时,产生超过20d B的附加路径损耗,并且随着我们的差距正在迅速增加。此外,由于频带变化情况,NLo S路径损耗大于Lo S损耗。因此,为了利用毫米波频带中较小的波长,28GHz频段是更好的选择。但是全数字28GHz系统的功耗和高成本的局限性使得它不太可能利用现有的半导体技术来实现,所以应该采用具有可控波束的混合数字模拟波束成形来使复杂度小型化。因此,由于社交网络的革命性爆炸和用户对高速无线通信的巨大需求,更高数据速率的需求正在迅速增加,毫米波频段可成为下一代的解决方案的无线通信的超越。选择28 GHz频段的基本目标是宽带无用和信道散布性。在较高频率的无线通信环境如毫米波,天线阵列成为重要的组成部分。因此,天线阵列的正常应用是多个接收机的同时传输。但是,毫米波系统中的主要问题是硬件限制,从而使经典低频带上的多接收机MIMO波束成形技术的实现变得困难。将各种数据流复用到各种接收机需要应用一些波束形成形式来产生发射的信号,并且对波束形成矩阵的条目具有优选的控制。但是,这种波束成形通常处于传统低频系统的基带阶段。不幸的是,组合信号组件的系统成本,复杂性和功率耗尽使得全数字基带波束形成器不利于当前的制造技术。此外,波束形成矩阵的设计通常基于几乎完美的信道状态信息,这在较小的波长系统中难以获得,这是因为当在系统中使用大量天线时需要巨大的训练开销,波束形成后的小信噪比(SNR)。因此,多接收机波束成形的新颖算法考虑了毫米波系统的硬件约束。并且需要开发低复杂度的毫米波系统。为了改善28 GHz频段的链路预算,我们开发了一种用于下行链路多接收机28GHz系统的双相低复杂度混合模数数字波束成形,而我们假定在接收机侧进行模拟信号合并。建议的算法可以归纳如下:波束形成器和组合器之间的一般耦合已经被使用,但是额外的挑战是使用不同约束的波束形成操作的决斗不同的域。因此,我们引用所提出的算法,我们必须参考将波束成形计算分为两个阶段的主要思想,在第一阶段,发射机的RF波束形成器和接收机的RF组合器被联合设计最大化每个接收机的所需信号功率,忽略接收机之间的干扰。在第二阶段,发射机的数字波束形成器被设计成处理多接收机干扰。因此,在第一阶段,发射机和每个单个接收机设计RF波束成形和组合向量,以最大化所引用接收机的期望信号功率,并忽略其他接收机的干扰。由于这是典型的单接收机RF波束成形设计问题,所以对于不需要显式信道估计并且具有较低训练开销的单接收机系统,改进的有效波束训练算法可用于设计射频波束形成组合向量。在第二阶段,有效信道将由发射机和接收机的系统号进行训练。每个有效信道向量的维数小于原始信道矩阵。这不是有效信道具有较大发射机阵列的算法的情况,因此每个接收机使用码本对其有效信道进行量化,量化信道向量的索引将以已知位数反馈给发射机。最后,发射机设计了基于量化信道的零强制数字波束形成器。由于窄波束形成和28GHz信道的稀疏性,有效的MIMO信道有望进行有条件的调整,这使得采用简单的多接收机数字波束成形策略,如能够实现接近最佳性能的零强制。在单接收器毫米波系统之前,对模拟和数字波束形成器的单独和联合设计进行了调查。他们考虑了单个接收机单流MIMO-OFDM系统,其中模拟和数字波束形成器被依次设计为对不同频率子载波之间的所需接收信号强度或频谱效率应用最大化。或者,模拟和数字波束形成器被联合设计以最大化单接收机系统的速率。在本论文中,已经考虑了不同的设置,也就是多接收机下行链路传输。因此,我们工作中混合模拟数字波束成形的目标与以前的工作不同,因为我们还需要管理多接收机干扰。这个解决方案使我们完全不同的分析。所提出的算法将发射机混合波束形成器和接收机的模拟组合器设计成具有反馈开销和小的训练。所提出的算法的性能分析已经采用两种情况,在单路径信道和具有大量发射机和接收机天线的多路径信道和具有两种类型的均匀阵列,均匀线性阵列(ULA)和均匀矩阵(URA)。波束形成向量模拟和数字已经从量化的码本中选择,因此与混合波束形成的性能相比,与仅模拟波束形成和所有数字块对角化算法相比,由于联合量化和视觉的速率损失的表征系统。为了能够分析混合波束形成,我们必须考虑到模拟和数字波束形成器之间的耦合,因此性能分析将会非常显着。因此,我们对提出的算法的性能进行了两个案例研究,即单路径信道的情况和假设大量天线的多路径信道的情况。这些情况是特别感兴趣的,因为28GHz信道可能是稀疏的,即仅存在少数路径,并且发射机和接收机都需要应用大的天线阵列以具有有效的接收功率。此外,这些特殊情况的分析将对本文中已经展示的更一般的设置中的所提出的算法的性能给出有用的见解。我们假设完全了解有效信道并假定RF射束向导的角度可以取连续值,分析所提出的算法的总和。本论文改进了低复杂度功能的双相混合模数数字波束成形算法,用于下行链路多接收机毫米波系统。所提出的算法在具有可用于发射机的阵列的已知尺寸和具有有限反馈的接收机信道之间的假设下更为通用。我们可以简要列出本论文的主要贡献如下:改进用于多接收机28GHz系统的混合发射机波束成形和接收机组合的算法。我们假设接收机仅使用模拟组合器,而在发射机处实现混合模拟数字波束形成器,其中在所提出的系统中已经使用RF链的数量等于接收机的数量或更少的接收机数量。所提出的算法的设计旨在减少反馈开销和训练,以获得更接近的结果给无约束的解决方案。在单路径信道的假设下分析混合波束成形算法性能,然后假设在发射机和接收机侧都具有大尺寸阵列几何形状的多路径信道,这被称为28 GHz系统的有利设计。混合量化码本的特征在于平均速率损失,与所有数字无约束算法和模拟波束成形解决方案相比,可以区分混合波束形成的大增益。在本论文中,混合波束成形包括模拟和数字组合处理,受组合信号硬件和全射频功率消耗的启发。所以安排如下:系统架构和渠道模型的描述在第2章。在第3章中,多接收机混合波束成形组合的总和速率计算问题已经通过训练和反馈开销的关联来形成。然后描述了所提出的低复杂度双相混合波束形成组合算法。第4章包括所提出的算法的性能分析,在具有非常大数量的天线的单路径信道和多路径信道中,假设连续移相器角度。第5章演示了两种不同分析案例研究中所提出的算法与模拟波束成形和块对角化算法全数字(无约束)的比较。并通过使用两种类型的天线阵列ULA和URA。所提出的混合波束成形算法表明,即使具有相对较小的训练和反馈开销也表现出良好的性能,我们确信我们感谢28 GHz信道的备用性质,并且已经在发射机和接收机中部署了大量的天线。在本论文中,已经提出了用于下行链路多接收机28GHz系统的双相低复杂度混合模数数字波束成形算法,其利用大量天线和28GHz频带的稀疏信道特性。对于两个案例研究,性能分析首先被考虑,当信道是单路径时,第二个是当通道是具有非常大数量的天线的多路径时被部署的。对于上述情况,我们演示了所提出的双相混合波束成形算法的渐近最优性,以及仅针对模拟波束成形和全数字无约束波束成形的增益。仿真结果表明,即使已经部署了大尺寸阵列,也需要多接收机28 GHz系统中的干扰管理。利用小信道的反馈,对联合模拟数字码本量化的速率损失平均值进行了数值模拟分析。这些结果表明混合波束形成的增益对于RF角度量化不是很敏感。对于数字波束形成层来说,为了在仅模拟波束成形上保持合理的波束成形增益,具有很好的量化是重要的。性能分析和仿真结果表明,所提出的模型对于仅模拟波束成形提供了更高的总和速率,并且通过相对小的码本几乎实现了所有数字系统的块对角化的相同方法。
[Abstract]:Currently, wireless services are saturated, almost all of the spectrum below 6GHz and the growing demand for high-speed wireless communications, especially with immersive multimedia applications are entering smart devices (such as smartphones, tablets, laptops, etc.). Space video traffic accounts for 61% of mobile traffic and is expected to come in the next few years. The richness of Full HD video is shared through social media, and the content of ultra high definition and 3D video will occur in the future, prompting researchers to study the millimeter wave (mm wave) band as a substitute for the classic band and to use wide bandwidth privileges as the fifth generation wireless communication system and its coming. An effective solution for demand. The millimeter wave band, especially the 28GHz, is still at the stage of research. As a solution to the next generation of wireless communications, a significant channel propagation characteristic in the millimeter wave band is the path loss of the huge transmission signal through the band. Previous studies have shown the channel of millimeter wave. The measurement results, in which the frequency difference between the band spectrum and the band spectrum is worse than the traditional cellular frequency band, has also shown the possibility of wireless communication based on the beamforming technology, which is easier to be realized through the small wavelength millimeter wave propagation for the Lo S and the NLo S transmitters and receivers. It has been proved that when we jump from 1.8GHz to 28GHz and then to the 60GHz band, the additional path loss over 20d B is generated and is increasing with our gap. In addition, the NLo S path loss is greater than the Lo S loss due to the frequency band variation. Therefore, the 28GHz band is a better choice for the smaller wavelengths in the millimeter wave band. However, the power and high cost limitations of the full digital 28GHz system make it unlikely to use the existing semiconductor technology, so a hybrid digital analog beamforming with a controllable beam should be used to miniaturize the complexity. Therefore, the revolutionary explosion of social networks and the huge demand for high-speed wireless communication by users The demand for higher data rates is increasing rapidly, and millimeter wave bands can become the surpassing of wireless communications in the next generation of solutions. The basic goal of selecting 28 GHz bands is broadband uselessness and channel dispersion. Antenna array becomes an important component in a high frequency wireless communication environment such as millimeter waves. Therefore, antenna arrays The normal application is the simultaneous transmission of multiple receivers. However, the main problem in the millimeter wave system is the hardware limitation, which makes it difficult to realize the multi receiver MIMO beamforming technology on the classic low frequency band. Moreover, the beamforming matrix has a preferred control. However, the beamforming is usually in the baseband stage of the traditional low frequency system. Unfortunately, the system cost, complexity and power depletion of the combined signal components make the full digital baseband beamformer detrimental to the current manufacturing technology. In addition, the design of the beamforming matrix It is usually based on almost perfect channel state information, which is difficult to obtain in a smaller wavelength system because it requires huge training overhead and small signal to noise ratio (SNR) after the use of a large number of antennas in the system. Therefore, a novel algorithm for multi receiver beamforming takes into account the hardware constraints of the millimeter wave system. And it is necessary to take into account the hardware constraints of the millimeter wave system. In order to improve the low complexity millimeter wave system. In order to improve the link budget of the 28 GHz band, we developed a dual phase low complexity hybrid analog digital beamforming for the downlink multi receiver 28GHz system. We assume that the analog signal is merged on the receiver side. The proposed algorithm can be summarized as follows: beamformer and The general coupling between the combiners has been used, but the additional challenge is to use the different domain of the duel of the beamforming operation with different constraints. Therefore, we refer to the proposed algorithm. We must refer to the main idea of dividing the beamforming calculation into two stages, in the first stage, the RF beamformer and the receiver of the transmitter. The RF combiner is jointly designed to maximize the required signal power for each receiver and ignore the interference between the receivers. In the second stage, the transmitter's digital beamformer is designed to deal with multiple receiver interference. Therefore, in the first stage, the transmitter and each individual receiver design the RF beamforming and combination vector to maximize the The desired signal power of the receiver is referenced and the interference of other receivers is ignored. Since this is a typical single receiver RF beamforming design problem, the improved effective beam training algorithm can be used to design the combination direction of the radio frequency beamforming for a single receiver system that does not need explicit channel estimation and has low training overhead. In the second stage, the effective channel will be trained by the system number of the transmitter and receiver. The dimension of each effective channel vector is less than the original channel matrix. This is not the case of an efficient channel with a larger array of transmitters. Therefore, each receiver uses the codebook to quantify its effective channel and quantify the channel vector cable. The lead will be fed back to the transmitter with known digits. Finally, the transmitter designs a zero forced digital beamformer based on a quantized channel. The effective MIMO channel is expected to be conditional due to the narrow beam forming and the sparsity of the 28GHz channel, which makes it possible to use a simple multi receiver digital beamforming strategy, such as to achieve close proximity. Zero coercion for optimal performance. Before a single receiver millimeter wave system, a single and joint design of analog and digital beamformers is investigated. They consider a single receiver single flow MIMO-OFDM system, in which the analog and digital beamformers are designed in turn to receive the required signal intensity between different frequency subcarriers. Or the application of spectral efficiency is maximized. Or, analog and digital beamformers are jointly designed to maximize the rate of single receiver systems. In this paper, different settings have been taken into account, that is, multi receiver downlink transmission. Therefore, the target of hybrid analog beamforming is different from previous work in our work. We also need to manage multi receiver interference. This solution makes us completely different. The proposed algorithm designs the analog combiner of the transmitter hybrid beamformer and the receiver into a feedback overhead and small training. The performance analysis of the proposed algorithm has been used in two cases, in the single path channel and in the tool. A multipath channel with a large number of transmitter and receiver antennas and two types of uniform arrays, uniform linear array (ULA) and homogeneous matrix (URA). Beamforming vector simulation and numbers have been selected from the quantized codebook, thus compared with the performance of mixed beamforming and diagonalization with only simulated beamforming and all digital blocks. In comparison with the representation system of joint quantization and rate loss of vision. In order to be able to analyze hybrid beamforming, we have to consider the coupling between analog and digital beamformers, so the performance analysis will be very significant. Therefore, we have two case studies on the proposed algorithm, that is, single path channel. Conditions and assumptions about the multipath channel of a large number of antennas. These are particularly interesting because the 28GHz channel may be sparse, that is, only a few paths exist, and both the transmitter and receiver need to apply large antenna arrays to have effective receiving power. In addition, the analysis of these special cases will be shown in this article. We give a useful insight into the performance of the proposed algorithm in more general settings. We assume that the effective channel is fully understood and the point of view of the RF beam wizard can be taken as a continuous value, and the sum of the proposed algorithms is analyzed. This paper improves the dual phase mixed modulus digital beamforming algorithm of low complexity function for the downlink chain. The proposed algorithm is more general under the assumption between the known size of the array that can be used for the transmitter and the receiver channel with limited feedback. We can briefly list the main contributions of this paper as follows: improving the beamforming and receiving of the hybrid transmitter for the multi receiver 28GHz system. We assume that the receiver uses an analog combiner only to implement a hybrid analog digital beamformer at the transmitter, in which the number of RF chains already used in the proposed system is equal to the number of receivers or fewer receivers. The proposed algorithm is designed to reduce feedback overhead and training to obtain the results. A more close result is given to an unconstrained solution. The performance of the hybrid beamforming algorithm is analyzed under the assumption of a single path channel, and then a multi-path channel with large size array geometry is assumed on the transmitter and receiver side, which is called a favorable design of the 28 GHz system. The characteristic of the mixed quantization codebook is the average rate loss. As compared with all digital unconstrained algorithms and analog beamforming solutions, the large gain of mixed beamforming can be distinguished. In this paper, hybrid beamforming includes analog and digital combination processing, inspired by the combined signal hardware and full radio frequency power consumption. In the second chapter, in the third chapter, the total rate calculation problem of multi receiver hybrid beamforming combination has been formed by the association of training and feedback overhead. Then the proposed low complexity biphasic hybrid beamforming algorithm is described. The fourth chapter includes the performance analysis of the proposed algorithm, in a very large number of antennas. In the path channel and the multipath channel, the fifth chapter demonstrates the comparison between the algorithm proposed in two different analysis case studies and the analog beamforming and block diagonalization algorithm (unconstrained). The hybrid beamforming algorithm proposed by the two types of antenna arrays ULA and URA. shows that The relatively small training and feedback overhead also shows good performance. We are convinced that we are grateful for the reserve properties of the 28 GHz channel and have already deployed a large number of antennas in the transmitter and receiver. In this paper, the dual phase low complexity mixed modulus number for the downlink multiple receiver 28GHz system has been proposed in this paper. A word beamforming algorithm, which takes advantage of the sparse channel characteristics of a large number of antennas and 28GHz bands. For two case studies, the performance analysis is first considered, when the channel is a single path and the second is deployed when the channel is a multipath with a very large number of antennas. For the above case, we demonstrate the proposed biphase mixing. The asymptotic optimality of the beamforming algorithm and only for the gain of analog beamforming and all digital unconstrained beamforming. Simulation results show that, even if a large size array has been deployed, interference management in a multi receiver 28 GHz system is also required. The results show that the gain of the mixed beamforming is not very sensitive to the RF angle quantization. For digital beamforming, it is important to have good quantization in order to maintain a reasonable beamforming gain in the only beamforming. The model provides a higher total sum rate for only simulated beamforming, and the same method of diagonalization of all digital systems is almost realized through relatively small codebooks.
【学位授予单位】：哈尔滨工业大学
【学位级别】：硕士
【学位授予年份】：2017
【分类号】：TN919.3

【相似文献】