WO2019041470A1 - 大规模mimo鲁棒预编码传输方法 - Google Patents

大规模mimo鲁棒预编码传输方法 Download PDF

Info

Publication number
WO2019041470A1
WO2019041470A1 PCT/CN2017/106351 CN2017106351W WO2019041470A1 WO 2019041470 A1 WO2019041470 A1 WO 2019041470A1 CN 2017106351 W CN2017106351 W CN 2017106351W WO 2019041470 A1 WO2019041470 A1 WO 2019041470A1
Authority
WO
WIPO (PCT)
Prior art keywords
channel
precoding
model
robust
transmission
Prior art date
Application number
PCT/CN2017/106351
Other languages
English (en)
French (fr)
Inventor
高西奇
卢安安
仲文
Original Assignee
东南大学
Priority date (The priority date is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the date listed.)
Filing date
Publication date
Application filed by 东南大学 filed Critical 东南大学
Priority to US16/643,560 priority Critical patent/US11177863B2/en
Publication of WO2019041470A1 publication Critical patent/WO2019041470A1/zh

Links

Images

Classifications

    • H—ELECTRICITY
    • H04—ELECTRIC COMMUNICATION TECHNIQUE
    • H04B—TRANSMISSION
    • H04B7/00—Radio transmission systems, i.e. using radiation field
    • H04B7/02—Diversity systems; Multi-antenna system, i.e. transmission or reception using multiple antennas
    • H04B7/04—Diversity systems; Multi-antenna system, i.e. transmission or reception using multiple antennas using two or more spaced independent antennas
    • H04B7/0413—MIMO systems
    • H—ELECTRICITY
    • H04—ELECTRIC COMMUNICATION TECHNIQUE
    • H04B—TRANSMISSION
    • H04B7/00—Radio transmission systems, i.e. using radiation field
    • H04B7/02—Diversity systems; Multi-antenna system, i.e. transmission or reception using multiple antennas
    • H04B7/04—Diversity systems; Multi-antenna system, i.e. transmission or reception using multiple antennas using two or more spaced independent antennas
    • H04B7/0413—MIMO systems
    • H04B7/0456—Selection of precoding matrices or codebooks, e.g. using matrices antenna weighting
    • H—ELECTRICITY
    • H04—ELECTRIC COMMUNICATION TECHNIQUE
    • H04B—TRANSMISSION
    • H04B17/00—Monitoring; Testing
    • H04B17/30—Monitoring; Testing of propagation channels
    • H04B17/391—Modelling the propagation channel
    • H04B17/3912—Simulation models, e.g. distribution of spectral power density or received signal strength indicator [RSSI] for a given geographic region
    • H—ELECTRICITY
    • H04—ELECTRIC COMMUNICATION TECHNIQUE
    • H04B—TRANSMISSION
    • H04B7/00—Radio transmission systems, i.e. using radiation field
    • H04B7/02—Diversity systems; Multi-antenna system, i.e. transmission or reception using multiple antennas
    • H04B7/04—Diversity systems; Multi-antenna system, i.e. transmission or reception using multiple antennas using two or more spaced independent antennas
    • H04B7/0413—MIMO systems
    • H04B7/0417—Feedback systems

Definitions

  • the invention belongs to the field of communication technologies and relates to a massive MIMO robust precoding transmission method.
  • MIMO multiple-input multiple-output
  • the large-scale MIMO precoding transmission method is related to the performance gain that can be achieved by large-scale MIMO.
  • Precoding can be divided into linear precoding and nonlinear precoding.
  • Nonlinear precoding can achieve approximate optimal performance, but its complexity is too high.
  • the basic research in the large-scale MIMO related literature is linear precoding, and most of them are downlink linear precoding for single antenna users. For a single-antenna user massive MIMO downlink, linear precoding can achieve approximately optimal performance when the number of transmit antennas is much larger than the number of receive antennas.
  • the linear precoding commonly used in the literature is matched filter (MF, Match Filter) precoding and regularized zero forcing (RZF) precoding.
  • the user terminal in the mobile communication system has adopted a multi-antenna device.
  • multi-antenna users will continue to exist. Therefore, the research on large-scale MIMO precoding design for multi-antenna users cannot be avoided.
  • Nonlinear precoding has a very high complexity and is currently not applicable to massive MIMO.
  • simple linear precoding such as MF precoding and RZF precoding, cannot achieve approximate optimal performance when the number of transmitting antennas is limited.
  • linear precoding to maximize weighted traversal and rate is considered.
  • the precoding design depends on the channel state information (CSI, Channel State Information) that the base station can obtain.
  • CSI Channel State Information
  • the widely used iterative weighted minimum mean square error (WMMSE, Weighted MMSE) precoding method in the traditional multi-user MIMO system can be directly extended to the massive MIMO system.
  • the method can converge to a local optimal solution that maximizes the weighting and rate optimization problem.
  • WMMSE weighted minimum mean square error
  • WMMSE weighted minimum mean square error
  • BDMA Wavelength Division Multiple Access
  • JSDM Joint Spatial Division and Multiplexing
  • each user channel information available at the base station needs to be modeled as a joint correlation model of known channel mean and variance. Since the CSI obtained by the base station has both channel mean and variance information, neither the BDMA transmission method nor the JSDM method can be applied.
  • the present invention discloses a large-scale MIMO robust precoding transmission method, which can solve the adaptability problem of large-scale MIMO technology in various typical scenarios.
  • the present invention provides the following technical solutions:
  • a massive MIMO robust precoding transmission method under imperfect channel state information comprising: based on a pilot signal and a priori statistical channel model, a base station or a transmitting device obtains a posteriori statistical channel model of a mobile terminal or a receiving device channel, including a channel mean Or expected value, and variance information; the base station or transmitting device performs robust precoding transmission using a posteriori statistical channel model including channel mean or expected value, and variance information.
  • a priori statistical correlation channel model is obtained by the following steps:
  • the base station or the transmitting device obtains through uplink channel sounding
  • the a priori statistical correlation channel model adopts one of the following models: a joint correlation channel model, a separation correlation model and a full correlation model.
  • the posterior statistical channel model is obtained by the following steps:
  • the base station or the transmitting device obtains channel information by channel estimation and prediction by using an uplink pilot signal and a priori joint correlation channel model;
  • Channel information is obtained based on channel estimation, prediction, and feedback by the mobile terminal or the receiving device using the downlink pilot signal and the a priori joint correlation channel model.
  • the channel mean or expected value and the variance information in the a posteriori statistical channel model include a channel posterior mean or a posteriori expectation value, and a posteriori variance information.
  • channel posterior mean or expected value, and the posterior variance information include:
  • conditional mean or conditional expectation value and conditional variance information of the mobile terminal or the receiving device under the condition of the received downlink pilot signal are provided.
  • the a posteriori statistical channel model is a posteriori statistical channel model including channel estimation error, channel aging, and spatial correlation effects.
  • the posterior statistical correlation channel model and the prior statistical channel model adopt one of the following models: a joint correlation channel model, a separation correlation model and a full correlation model.
  • the base station or the transmitting device performs a linear precoding matrix design of each mobile terminal or receiving device according to the weighted traversal and rate maximization criteria, and the weighted traversal and rate are based on the established posterior The weighted sum rate conditional mean calculated by the statistical channel model.
  • the weighted traversal is performed by the MM algorithm.
  • the rate maximization precoding design problem is transformed into an iterative solution quadratic optimization problem.
  • the random matrix expectation required for solving the quadratic optimization problem is calculated by using the deterministic equivalent.
  • a method for acquiring a large-scale MIMO downlink robust precoding domain pilot multiplex channel information under the imperfect channel state information includes: obtaining, by the base station or the transmitting device, a posterior statistical channel model of the mobile terminal or the receiving device channel, including a channel mean or an expected value, And the variance information; the base station or the transmitting device performs robust precoding transmission using a posteriori statistical channel model including channel mean or expected value and variance information; in the robust precoding transmission, the downlink is implemented in the precoding domain
  • the pilot multiplex channel information is acquired, and the base station or the transmitting device sends a downlink pilot signal to each mobile terminal or the receiving device in the precoding domain, and the mobile terminal or the receiving device uses the received pilot signal to perform the precoding domain equivalent channel.
  • Channel estimation, the precoding domain equivalent channel is the actual transmission channel multiplied by the robust precoding matrix.
  • the precoding domain pilot signals transmitted by the base station or the transmitting device to each mobile terminal or the receiving device are transmitted on the same time-frequency resource, and the pilots of the mobile terminals or the receiving devices are not required to be orthogonal.
  • the precoding domain pilot signal transmitted by the base station or the transmitting device to each mobile terminal or the receiving device is a frequency domain signal modulated by the ZC sequence or the ZC sequence group.
  • a receiving method for massive MIMO robust precoding transmission includes: a transmitting signal transmitted by robust precoding is received by a mobile terminal or a receiving device after passing through a transmission channel, and the mobile terminal or the receiving device uses the received transmitting signal to receive a signal. deal with.
  • the received transmission signal includes a downlink full pilot frequency signal, and/or a robust precoding domain pilot signal, and/or a robust precoding domain data signal.
  • the pilot signal is a frequency domain signal generated by modulation of a ZC sequence or a ZC sequence group.
  • the mobile terminal or the receiving device performs channel estimation, prediction, and feedback using the received downlink full pilot frequency signal.
  • the mobile terminal or the receiving device performs channel estimation of the precoding domain equivalent channel by using the received robust precoding domain pilot signal.
  • the mobile terminal or the receiving device performs demodulation or detection of the precoding domain signal by using the received robust precoding domain data signal.
  • the present invention has the following advantages and benefits:
  • the large-scale MIMO robust transmission method proposed by the present invention can solve the universality problem of large-scale MIMO for various typical mobile scenarios and achieve high spectral efficiency.
  • the invention utilizes a posteriori statistical channel model including channel mean and variance information for robust precoding transmission, and the statistical channel model used is more complete and accurate; the robust precoding method can realize dimensionality reduction transmission, which can reduce the data transmission required
  • the pilot overhead reduces the complexity of demodulation or detection and improves the overall efficiency of the transmission.
  • 1 is a flowchart of a massive MIMO downlink robust precoding transmission method
  • 2 is a flowchart of another method for massive MIMO downlink robust precoding transmission
  • FIG. 3 is a flowchart of a method for acquiring channel information of a massive MIMO downlink robust precoding domain
  • FIG. 4 is a flowchart of a receiving method for massive MIMO downlink robust precoding transmission
  • 5 is a flow chart of another receiving method for massive MIMO downlink robust precoding transmission
  • FIG. 6 is a flowchart of a massive MIMO downlink robust linear precoding design method
  • FIG. 7 is a schematic diagram of comparison of traversal and rate results of the robust precoding transmission method and the BDMA method in the present embodiment under the downlink of the massive MIMO system;
  • FIG. 8 is a schematic diagram of a robust precoding transmission method in this embodiment under the downlink of a massive MIMO system under consideration. A comparison of the traversal and rate results of the rod RZF method in three different mobile scenarios.
  • a method for transmitting a large-scale MIMO downlink robust precoding includes: a base station obtains a prior channel joint correlation channel model of each user channel by using an uplink channel sounding; and the base station uses an uplink pilot signal and a first A joint correlation channel model is obtained, and a posteriori joint correlation channel model of each user channel is obtained through channel estimation and prediction, including channel mean and variance information; the base station performs robust precoding using a posteriori joint correlation channel model including channel mean and variance information. transmission.
  • the base station referred to in the present invention may also employ other transmitting devices capable of transmitting and transmitting information.
  • the channel mean is also often referred to as the expected value.
  • another large-scale MIMO downlink robust precoding transmission method disclosed in the embodiment of the present invention includes a mobile terminal obtaining a channel prior a joint correlation channel model through downlink channel sounding; and a mobile terminal using a downlink pilot signal. And a priori joint channel model, obtaining a posteriori joint correlation channel model of each channel through channel estimation and prediction and feeding back to the base station, the posterior joint correlation model including channel mean and variance information; the base station utilizing channel mean and variance information The posterior joint correlation model performs robust precoding transmission.
  • the mobile device referred to in the present invention may also employ other receiving devices capable of receiving information.
  • a method for acquiring a large-scale MIMO downlink robust precoding domain pilot multiplexed channel information includes a base station using a pilot signal and a prior statistical channel model to obtain a posterior of a mobile terminal.
  • the statistical channel model includes channel mean and variance information; the base station performs robust precoding design using a posteriori statistical channel model including channel mean and variance information; the base station transmits robust precoding domain pilots to each user on the same time-frequency resource
  • the signal, the robust precoding domain pilot signal used is a ZC sequence; the mobile terminal uses the received robust precoding domain pilot signal to perform channel estimation of the robust precoding domain equivalent channel, and the robust precoding domain equivalent
  • the channel is the actual transmission channel multiplied by the robust precoding matrix.
  • a receiving method for massive MIMO downlink robust precoding transmission disclosed in the embodiment of the present invention includes: using a pilot signal and a prior statistical channel model to obtain a posteriori statistical channel model of a mobile terminal, including Channel mean and variance information; the base station or the transmitting device performs robust precoding design using a posteriori joint correlation channel model including channel mean and variance information; the base station transmits the robust precoded signal, including the robust precoding domain guide Frequency signals and data signals; mobile terminals use the received robust precoding domain pilot signals for robustness Channel estimation of the precoding domain equivalent channel; the mobile terminal performs precoding domain signal demodulation or detection using the received data signal and the channel estimation of the precoding domain equivalent channel.
  • another method for receiving massive MIMO downlink robust precoding transmission includes: the base station sends a downlink full pilot frequency; and the mobile terminal uses the received full pilot frequency for channel estimation and prediction. And a posteriori statistical channel model of the feedback mobile terminal, including channel mean and variance information; the base station or the transmitting device performs robust precoding design using a posteriori joint correlation channel model including channel mean and variance information; the base station transmits the robust precoding
  • the processed signal includes a robust precoding domain pilot signal and a data signal; the mobile terminal utilizes the received robust precoding domain pilot signal to perform channel estimation of the robust precoding domain equivalent channel; and the mobile terminal utilizes the receiving The pre-coded domain signal demodulation or detection is performed on the received data signal and the channel estimate of the precoding domain equivalent channel.
  • a method for designing a massive MIMO downlink robust precoding includes a base station establishing a weighted traversal and a rate maximization problem according to a posterior statistical channel model; and a weighted traversal and rate by an MM algorithm.
  • the maximization problem is transformed into an iterative solution to the quadratic problem; the random matrix expectation of the closed-type optimal solution of the quadratic problem is used, and its deterministic equivalence is used for fast calculation.
  • the method of the present invention is mainly applicable to a large-scale MIMO system equipped with a large-scale antenna array on the base station side to simultaneously serve multiple users.
  • the specific implementation process of the robust precoding transmission method of the present invention is described in detail below with reference to a specific communication system example. It should be noted that the method of the present invention is applicable not only to the specific system model cited in the following examples, but also to other configurations. System model.
  • the MIMO system consists of one base station and K mobile terminals.
  • the number of antennas configured by the base station is M t .
  • the number of antennas configured by the kth user is M k , and
  • the system time resource is divided into a number of time slots, each of which includes N b time blocks (T symbol intervals).
  • the large-scale MIMO system considered in this embodiment operates in a Time Division Duplexing (TDD) mode.
  • TDD Time Division Duplexing
  • the uplink pilot signal is transmitted only in the first block.
  • the second to N b blocks are used for downlink precoding domain pilot signals and data signal transmission.
  • the length of the uplink training sequence is the length of the block, that is, T symbol intervals.
  • mutually orthogonal training sequences (M r ⁇ T) are used for different uplink transmit antennas.
  • FDD Frequency Division Duplexing
  • the uplink channel training phase can be replaced with the downlink channel feedback phase, and the downlink transmission phase remains unchanged. Specifically, the downlink full pilot signal is transmitted in the first block, and the mobile terminal feedback is received.
  • the large-scale MIMO system channel is considered to be a stationary channel, and the statistical channel model of each user channel can be represented as a joint correlation model.
  • the channel from the base station to the kth user on the nth block of the mth slot has the following structure
  • U k and V k are deterministic ⁇ matrices
  • M k is a deterministic matrix composed of non-negative elements
  • W k,m,n is a matrix consisting of zero mean, unit variance, independent and identically distributed complex Gaussian random variables .
  • M t can become very large.
  • all users have the same V k .
  • the base station side is equipped with a uniform linear antenna array and the number of antennas M t is very large.
  • all users' V k can be approximated as a DFT matrix.
  • the channel model in equation (1) can be rewritten as
  • Equation (2) Represents an M t ⁇ M t -dimensional DFT matrix.
  • the channel model described in equation (2) can be regarded as a priori model of the channel before channel estimation.
  • the time evolution of the channel between blocks and blocks is represented by a first-order Gauss-Markov stochastic model as
  • ⁇ k is a time correlation factor related to the user's moving speed.
  • J 0 ( ⁇ ) represents the first-order zero-order Bessel function
  • vk represents the kth User speed
  • f c represents the carrier frequency
  • c is the speed of light.
  • the model in equation (3) is used for channel prediction.
  • the base station obtains a priori joint correlation channel model of each user through channel sounding, that is, obtains U k and ⁇ k .
  • a priori joint correlation channel model of each user can be obtained through user downlink channel sounding.
  • the CSI of the downlink transmission channel is obtained by channel estimation using the uplink pilot signal received by the base station through channel reciprocity.
  • Uplink pilot matrix of the kth user in the first block on the time slot m Indicates an uplink received random noise matrix whose elements have a mean of zero and a variance of And independent and identically distributed complex Gaussian random variables.
  • W k,m,n is a matrix consisting of independent and identically distributed, zero mean, unit variance complex Gaussian random variable elements, Medium element is
  • Equation (7) indicates that the imperfect CSI of each UE available at the base station side can be modeled as a joint correlation model including channel mean (or expected value) and variance information, and the model includes channel estimation error, channel variation, and Spatially related impacts.
  • the channel information obtained by the base station in the equation (7) is a conditional mean value (or conditional expectation value) and conditional variance information of the base station under the condition that the uplink pilot signal is received.
  • the a posteriori model described in the equation (7) is a general model of the imperfect CSI available at the base station side of the massive MIMO system in different mobile scenarios. When ⁇ k is very close to 1, it is suitable for a communication scenario when the user is quasi-stationary.
  • ⁇ k becomes very small, it is suitable for a communication scenario in which the user moves very fast. Further, in this case It becomes close to zero, and the difference between the a posteriori model in the equation (7) and the a priori model in the equation (2) becomes very small.
  • the ⁇ k is set to different values according to different moving speeds of the user, and the established a posteriori model can be used to describe the channel model in a plurality of typical mobile communication scenarios of massive MIMO.
  • the channel estimation, prediction and feedback of the mobile terminal can also obtain the posterior joint correlation model in equation (7).
  • the base station transmits a downlink full pilot frequency; the mobile terminal performs channel estimation, prediction, and feedback using the received full pilot frequency.
  • the channel information obtained in the equation (7) becomes a conditional mean value (or conditional expectation value) and conditional variance information of the mobile terminal under the condition of the received downlink pilot signal.
  • P k,m,n are the M k ⁇ d k -dimensional precoding matrices of the kth UE
  • z k,m,n is a distribution Complex Gaussian random noise vector
  • the transmitted robust precoding domain pilot signals are on the same time-frequency resource, and each user pilot does not require orthogonality, that is, pilot multiplexing can be performed.
  • the precoding domain pilot signal transmitted by the base station to each user is a frequency domain signal generated by modulation of a ZC sequence or a ZC sequence group.
  • the mobile terminal After receiving the pilot signal, the mobile terminal performs channel estimation of the robust precoding domain equivalent channel, and the robust precoding domain equivalent channel is H k,m,n P k,m,n .
  • the UE side can obtain a perfect CSI with an equivalent channel matrix of the respective robust precoding domain.
  • each user can perform robust precoding domain signal detection by using the received data signal. Total interference noise per UE
  • R k,m,n denote the covariance matrix of z' k,m,n .
  • Expectation function ⁇ represents the expected function of H k,m,n based on the user side length statistics. According to the channel reciprocity, the long-term statistical channel information of the user side is consistent with the long-term statistical channel information of the base station end given in the formula (2). Therefore, the expectation function ⁇ can be calculated according to equation (2). Suppose the kth UE knows R k,m,n , and the kth user traversal rate can be expressed as
  • a conditional expectation function for H k,m,n derived from the posterior model in equation (7) is represented.
  • Defining function represents the weighted traversal and rate, which is the weighted sum rate conditional mean calculated from the established posterior statistical channel model.
  • the purpose of this embodiment is to design the precoding matrix P 1,m,n ,P 2,m,n ,...,P K,m,n to maximize the weighted traversal and rate, ie to solve the optimization problem.
  • w k is the weighting factor of the kth user and P is the total power constraint.
  • the objective function in the optimization problem (13) is a very complex function about the precoding matrix, so the problem is difficult to solve directly.
  • the MM (Minorize Maximization or Majorize Minimization) algorithm can be used to transform the weighted traversal and rate maximization precoding design problem into an iterative solution quadratic optimization problem.
  • the key to the MM algorithm is to find a simple minorizing function for the objective function. For simplicity, define a single-sided correlation matrix with for
  • the limit point of the precoding matrix sequence given in equation (26) is a local maximum point of the original optimization problem (13). Further, the optimization problem in the equation (26) is a concave quadratic function of the precoding matrix P 1, m, n , P 2, m, n , ..., P K, m, n .
  • the optimal solution can be obtained directly by the Lagrange multiplier method.
  • ⁇ * is the optimal Lagrangian factor corresponding to the energy constraint.
  • the channel model given in equation (7) is a joint correlation model with a non-zero mean.
  • the deterministic equivalent of R k,m,n is
  • Step 3 Calculate ⁇ k, m, n and according to equations (34)-(35)
  • Step 4 Calculate according to equations (22), (38), (39) with
  • a comparison of the results of the robust precoding transmission method and the BDMA method in the present embodiment is given.
  • Figure 7 shows the traversal and rate results comparison of the robust precoding transmission method and the BDMA method in the present embodiment under the downlink of the massive MIMO system under consideration.
  • the robust precoding transmission method in this embodiment is significantly superior to the BDMA method. This is because in the robust precoding transmission method, the base station performs robust precoding transmission using a posteriori joint correlation model including channel mean (or expected value) and variance information, and the BDMA method utilizes prior joint correlation including only channel variance information. The model transmits and fails to make full use of the statistical channel information available to the base station.
  • the robust RZF method is an extension of the RZF precoding method widely used in single-antenna user massive MIMO systems under imperfect channel state information.

Landscapes

  • Engineering & Computer Science (AREA)
  • Computer Networks & Wireless Communication (AREA)
  • Signal Processing (AREA)
  • Physics & Mathematics (AREA)
  • Spectroscopy & Molecular Physics (AREA)
  • Electromagnetism (AREA)
  • Mobile Radio Communication Systems (AREA)
  • Radio Transmission System (AREA)

Abstract

本发明公开了一种非完美信道状态信息下大规模MIMO鲁棒预编码传输方法。所公开方法中，大规模MIMO***基站端获得的非完美信道状态信息被建模为包含信道均值和方差信息的后验统计信道模型。该模型考虑了信道估计误差、信道老化以及空间相关的影响。本发明中，基站利用后验统计信道模型进行鲁棒预编码传输，能够解决大规模MIMO对各种典型移动场景的普适性问题，并取得高频谱效率。本发明利用包括信道均值和方差信息的后验统计信道模型进行鲁棒预编码传输，所用统计信道模型更加充分和准确；通过鲁棒预编码方法可以实现降维传输，可以降低数据传输时所需的导频开销，降低解调或检测的复杂度，提高传输的整体效率。

Description

大规模MIMO鲁棒预编码传输方法 技术领域
本发明属于通信技术领域，涉及大规模MIMO鲁棒预编码传输方法。
背景技术
为提升用户体验，应对无线数据业务需求的快速增长以及新业务需求带来的挑战，未来新一代移动网络需要支持高质量、高传输率、高移动性、高用户密度、低时延等场景。通过在基站(BS，BaseStation)配备大规模天线阵列极大的提高***容量的大规模多输入多输出(MIMO，Multiple-Input Multiple-Ouput)技术，是未来新一代无线网络的关键技术之一，也是近年来的研究热点。
大规模MIMO预编码传输方法关系到能否实现大规模MIMO所能提供的性能增益。预编码可分为线性预编码和非线性预编码。非线性预编码能够取得近似最优性能，但是其复杂度太高。大规模MIMO相关文献中目前研究的基本都是线性预编码，并且大部分都是针对单天线用户的下行线性预编码。对于单天线用户大规模MIMO下行链路，在发送天线数量远大于接收天线数量时，线性预编码可取得近似最优的性能。文献中常用的线性预编码有匹配滤波(MF，Match Filter)预编码和正则化迫零(RZF，Regularized Zero Forcing)预编码。
目前移动通信***中用户端已经采用多天线设备。在未来新一代无线网络中，多天线用户必然继续存在。因此，针对多天线用户大规模MIMO预编码设计的研究无法回避。非线性预编码复杂度极高，目前还无法应用到大规模MIMO。同时，简单的线性预编码，如MF预编码和RZF预编码等，在发送天线数量有限的情况下，无法取得近似最优性能。为取得接近最优性能，需考虑最大化加权遍历和速率的线性预编码。预编码设计取决于基站能获得的信道状态信息(CSI，Channel State Information)。当基站侧具有完美CSI时，传统多用户MIMO***中广泛使用的迭代加权最小均方误差(WMMSE，Weighted MMSE)预编码方法可直接推广到大规模MIMO***。该方法可收敛到最大化加权和速率优化问题的局部最优解。当基站具有统计CSI时，文献中有波分多址(BDMA，Beam Division Multiple Access)传输方法和联合空分复用(JSDM，Joint Spatial Division and Multiplexing)方法。具体而言，这两种方法都工作于统计CSI具有零均值 时的情形。
实际大规模MIMO***中，由于信道估计误差、信道变化以及其他因素的影响，通常基站端不能获得完美CSI。因此，实际***中，基于完美CSI的迭代WMMSE算法无法使用。而为了满足实际***中各用户可能处于不同移动场景需求，需要将基站端可获得的各用户信道信息建模为已知信道均值和方差的联合相关模型。由于基站所获得的CSI同时具有信道均值和方差信息，BDMA传输方法和JSDM方法也都不能应用。
发明内容
针对现有技术的不足，本发明公开了大规模MIMO鲁棒预编码传输方法，能够解决大规模MIMO技术在各种典型场景下的适应性问题。
为了达到上述目的，本发明提供如下技术方案：
非完美信道状态信息下大规模MIMO鲁棒预编码传输方法，包括：基于导频信号和先验统计信道模型，基站或发送装置获得移动终端或接收装置信道的后验统计信道模型，包含信道均值或期望值、以及方差信息；基站或发送装置利用包含信道均值或期望值、以及方差信息的后验统计信道模型进行鲁棒预编码传输。
进一步，所述先验统计相关信道模型通过以下步骤获得：
基站或发送装置通过上行信道探测获得；
或
通过移动终端或接收装置基于下行信道探测获得。
进一步，所述先验统计相关信道模型采用以下模型中的一种：联合相关信道模型，分离相关模型和全相关模型。
进一步，所述后验统计信道模型通过以下步骤获得：
基站或发送装置利用上行导频信号和先验联合相关信道模型，通过信道估计和预测获得信道信息；
或
通过移动终端或接收装置利用下行导频信号和先验联合相关信道模型，基于信道估计、预测、反馈获得信道信息。
进一步，所述后验统计信道模型中信道均值或期望值、以及方差信息包括信道后验均值或后验期望值、以及后验方差信息。
进一步，所述信道后验均值或期望值、以及后验方差信息包括：
基站或发送装置在接收到的上行导频信号条件下的条件均值或条件期望值、以及条件方差信息；
或
移动终端或接收装置在接收到的下行导频信号条件下的条件均值或条件期望值、以及条件方差信息。
进一步，所述后验统计信道模型为包含信道估计误差、信道老化和空间相关影响的后验统计信道模型。
进一步，所述后验统计相关信道模型和先验统计信道模型采用以下模型中的一种：联合相关信道模型，分离相关模型和全相关模型。
进一步，在所述的鲁棒预编码传输中，基站或发送装置根据加权遍历和速率最大化准则，进行各移动终端或接收装置的线性预编码矩阵设计，加权遍历和速率为根据所建立后验统计信道模型计算出的加权和速率条件均值。
进一步，在所述的鲁棒预编码传输过程中，基站或发送装置根据加权遍历和速率最大化准则进行各移动终端或接收装置的线性预编码设计时，通过MM算法将所述将加权遍历和速率最大化预编码设计问题转化为迭代求解二次型优化问题进行求解。
进一步，所述二次型优化问题求解时所需的随机矩阵期望，利用其确定性等同进行快速计算。
非完美信道状态信息下大规模MIMO下行鲁棒预编码域导频复用信道信息获取方法，包括：基站或发送装置获得移动终端或接收装置信道的后验统计信道模型，包含信道均值或期望值、以及方差信息；基站或发送装置利用包含信道均值或期望值、以及方差信息的后验统计信道模型进行鲁棒预编码传输；在所述的鲁棒预编码传输中，下行链路在预编码域实施导频复用信道信息获取，基站或发送装置在预编码域向各移动终端或接收装置发送下行导频信号，移动终端或接收装置利用接收到的导频信号，进行预编码域等效信道的信道估计，预编码域等效信道为实际的传输信道乘以鲁棒预编码矩阵。
进一步，基站或发送装置向各移动终端或接收装置发送的预编码域导频信号在同一时频资源上发送，各移动终端或接收装置的导频不要求正交。
进一步，基站或发送装置向各移动终端或接收装置发送的预编码域导频信号为ZC序列或ZC序列组经过调制生成的频域信号。
大规模MIMO鲁棒预编码传输的接收方法，包括：通过鲁棒预编码传输的发送信号经过传输信道后由移动终端或接收装置进行接收，移动终端或者接收装置利用接收到的发送信号进行接收信号处理。
进一步，所述的接收到的发送信号包括下行全向导频信号，和/或鲁棒预编码域导频信号，和/或鲁棒预编码域数据信号。
进一步，所述导频信号为ZC序列或ZC序列组经过调制生成的频域信号。
进一步，在所述的接收信号处理中，移动终端或接收装置利用接收到的下行全向导频信号进行信道估计、预测和反馈。
进一步，在所述的接收信号处理中，移动终端或接收装置利用接收到的鲁棒预编码域导频信号进行预编码域等效信道的信道估计。
进一步，在所述的接收信号处理中，移动终端或接收装置利用接收到的鲁棒预编码域数据信号进行预编码域信号的解调或检测。
与现有技术相比，本发明具有如下优点和有益效果：
本发明提出的大规模MIMO鲁棒传输方法能够解决大规模MIMO对各种典型移动场景的普适性问题，并取得高频谱效率。本发明利用包括信道均值和方差信息的后验统计信道模型进行鲁棒预编码传输，所用统计信道模型更加充分和准确；通过鲁棒预编码方法可以实现降维传输，可以降低数据传输时所需的导频开销，降低解调或检测的复杂度，提高传输的整体效率。
附图说明
图1为一种大规模MIMO下行鲁棒预编码传输方法流程图；
图2为另一种大规模MIMO下行鲁棒预编码传输方法流程图；
图3为一种大规模MIMO下行鲁棒预编码域信道信息获取方法流程图；
图4为一种大规模MIMO下行鲁棒预编码传输的接收方法流程图；
图5为另一种大规模MIMO下行鲁棒预编码传输的接收方法流程图；
图6为一种大规模MIMO下行鲁棒线性预编码设计方法流程图；
图7为在所考虑大规模MIMO***下行链路下，本实施例中鲁棒预编码传输方法与BDMA方法的遍历和速率结果比较示意图；
图8为在所考虑大规模MIMO***下行链路下，本实施例中鲁棒预编码传输方法与鲁 棒RZF方法在三种不同移动场景下的遍历和速率结果比较示意图。
具体实施方式
以下将结合具体实施例对本发明提供的技术方案进行详细说明，应理解下述具体实施方式仅用于说明本发明而不用于限制本发明的范围。
如图1所示，本发明实施例公开的一种大规模MIMO下行鲁棒预编码传输方法，包括基站通过上行信道探测获得各用户信道先验联合相关信道模型；基站利用上行导频信号和先验联合相关信道模型，通过信道估计和预测获得各用户信道的后验联合相关信道模型，包含信道均值和方差信息；基站利用包含信道均值和方差信息的后验联合相关信道模型进行鲁棒预编码传输。本发明所指的基站也可以采用其他能够进行信息发送、传输的发送装置。在本领域内，信道均值也常被称为期望值。
如图2所示，本发明实施例公开的另一种大规模MIMO下行鲁棒预编码传输方法，包括移动终端通过下行信道探测获得各自信道先验联合相关信道模型；移动终端利用下行导频信号和先验联合相关信道模型，通过信道估计和预测获得各自信道的后验联合相关信道模型并反馈给基站，该后验联合相关模型包含信道均值和方差信息；基站利用包含信道均值和方差信息的后验联合相关模型进行鲁棒预编码传输。本发明所指的移动也可以采用其他能够进行信息接收的接收装置。
如图3所示，本发明实施例公开的一种大规模MIMO下行鲁棒预编码域导频复用信道信息获取方法，包括基站利用导频信号和先验统计信道模型获得移动终端的后验统计信道模型，包含信道均值和方差信息；基站利用包含信道均值和方差信息的后验统计信道模型进行鲁棒预编码设计；基站在同一时频资源上向各用户发送鲁棒预编码域导频信号，所用鲁棒预编码域导频信号为ZC序列；移动终端利用接收到的鲁棒预编码域导频信号，进行鲁棒预编码域等效信道的信道估计，鲁棒预编码域等效信道为实际的传输信道乘以鲁棒预编码矩阵。
如图4所示，本发明实施例公开的一种大规模MIMO下行鲁棒预编码传输的接收方法，包括基站利用导频信号和先验统计信道模型获得移动终端的后验统计信道模型，包含信道均值和方差信息；基站或发送装置利用包含信道均值和方差信息的后验联合相关信道模型进行鲁棒预编码设计；基站发送经过鲁棒预编码处理过的信号，包括鲁棒预编码域导频信号和数据信号；移动终端利用接收到的鲁棒预编码域导频信号，进行鲁棒 预编码域等效信道的信道估计；移动终端利用接收到的数据信号和预编码域等效信道的信道估计进行预编码域信号解调或检测。
如图5所示，本发明实施例公开的另一种大规模MIMO下行鲁棒预编码传输的接收方法，包括基站发送下行全向导频；移动终端利用接收到的全向导频进行信道估计、预测与反馈移动终端的后验统计信道模型，包含信道均值和方差信息；基站或发送装置利用包含信道均值和方差信息的后验联合相关信道模型进行鲁棒预编码设计；基站发送经过鲁棒预编码处理过的信号，包括鲁棒预编码域导频信号和数据信号；移动终端利用接收到的鲁棒预编码域导频信号，进行鲁棒预编码域等效信道的信道估计；移动终端利用接收到的数据信号和预编码域等效信道的信道估计进行预编码域信号解调或检测。
如图6所示，本发明实施例公开的一种大规模MIMO下行鲁棒预编码设计方法，包括基站根据后验统计信道模型建立加权遍历和速率最大化问题；通过MM算法将加权遍历和速率最大化问题转化为迭代求解二次型问题；将二次型问题闭式最优解所需随机矩阵期望，使用其确定性等同进行快速计算。
本发明方法主要适用于基站侧配备大规模天线阵列以同时服务多个用户的大规模MIMO***。下面结合具体的通信***实例对本发明涉及鲁棒预编码传输方法的具体实现过程作详细说明，需要说明的是本发明方法不仅适用于下面示例所举的具体***模型，也同样适用于其它配置的***模型。
一、***配置
考虑一平坦块衰落大规模MIMO***，其各用户信道在T个符号间隔内保持不变。该MIMO***由一个基站和K个移动终端构成。基站配置的天线数量为Mt。第k用户配置的天线数为Mk，并且
Figure PCTCN2017106351-appb-000001
将***时间资源分为若干个时隙，每一时隙包括Nb个时间块(T个符号间隔)。本实施例中所考虑大规模MIMO***工作于时分双工(TDD，Time Division Duplexing)模式。简便起见，假设只存在上行信道训练阶段和下行传输阶段，下行传输阶段包括预编码域导频信号发送和预编码域数据信号发送。在每一时隙中，只在第一块传输上行导频信号。第2至Nb块则用于下行预编码域导频信号和数据信号传输。上行训练序列的长度为块的长度，即T个符号间隔。进一步，对于不同上行发送天线使用互相正交的训练序列(Mr≤T)。对于频分双工(FDD，Frequency Division Duplexing)模式，可以将上行信道训练阶段替换为下行信道反馈阶段，下行传 输阶段则保持不变。具体来说，在第一块传输下行全向导频信号，并接收移动终端反馈。二、先验统计信道模型
假设所考虑大规模MIMO***信道为平稳信道，并且每一用户信道的统计信道模型可以表示为联合相关模型。具体来说，在第m时隙第n块上基站到第k用户的信道有如下结构
Figure PCTCN2017106351-appb-000002
其中，Uk和Vk为确定酉矩阵，Mk为一由非负元素组成的确定矩阵，Wk,m,n为一由零均值、单位方差、独立同分布复高斯随机变量组成的矩阵。在大规模MIMO***里，Mt可以变得非常大。此时，所有用户的Vk相同。本实施例假设基站端配备均匀线性天线阵列并且天线数量Mt非常大。在这种场景下，所有用户的Vk可以近似为一DFT矩阵。综上所述，式(1)中的信道模型可以重新写为
Figure PCTCN2017106351-appb-000003
其中
Figure PCTCN2017106351-appb-000004
表示一Mt×Mt维DFT矩阵。式(2)中所述信道模型可以看作是信道估计前信道的先验(a priori)模型。进一步，将信道在块与块之间的时间演变用一阶Gauss-Markov随机模型表示为
Figure PCTCN2017106351-appb-000005
其中αk是和用户移动速度有关的时间相关因子。采用常用的基于Jakes自相关模型的αk的计算方法，即αk＝J0(2πvkfcT/c)，其中J0(·)表示第一类零阶Bessel函数，vk表示第k用户速度，fc表示载波频率，c为光速。式(3)中模型用来进行信道预测。
定义大规模MIMO***信道能量耦合矩阵Ωk为Ωk＝Mk⊙Mk。对于所考虑工作于TDD模式的大规模MIMO***，假设基站通过上行信道探测(channel sounding)获得各用户先验联合相关信道模型，即获得Uk和Ωk。对于工作于FDD模式的大规模MIMO***，则可以通过用户下行信道探测获得各用户先验联合相关信道模型。
三、后验统计信道模型
对于所考虑工作于TDD模式的大规模MIMO***，通过信道互易性，利用基站接收到的上行导频信号，进行信道估计获得下行传输信道的CSI。令
Figure PCTCN2017106351-appb-000006
表示在时隙m 第一块上基站的接收矩阵，
Figure PCTCN2017106351-appb-000007
表示Mt×T复矩阵集。可以将其写为
Figure PCTCN2017106351-appb-000008
其中
Figure PCTCN2017106351-appb-000009
表示在时隙m上第一块内第k用户的上行导频矩阵，
Figure PCTCN2017106351-appb-000010
表示上行接收随机噪声矩阵，其元素为均值为零、方差为
Figure PCTCN2017106351-appb-000011
且独立同分布的复高斯随机变量。
在给定
Figure PCTCN2017106351-appb-000012
时，可得Hk,m,n的后验(a posteriori)均值，即MMSE估计
Figure PCTCN2017106351-appb-000013
为
Figure PCTCN2017106351-appb-000014
其中
Figure PCTCN2017106351-appb-000015
进一步，可得给定
Figure PCTCN2017106351-appb-000016
时Hk,m,n的后验模型为
Figure PCTCN2017106351-appb-000017
其中，Wk,m,n为一由独立同分布、零均值、单位方差复高斯随机变量元素组成的矩阵，
Figure PCTCN2017106351-appb-000018
中元素为
Figure PCTCN2017106351-appb-000019
式(7)表示，基站侧可获得的每一UE的非完美CSI可以建模为包含信道均值(或称期望值)和方差信息的联合相关模型，并且该模型包含了信道估计误差、信道变化和空间相关的影响。式(7)中基站所获得的信道信息为基站在接收到上行导频信号条件下的条件均值(或称条件期望值)和条件方差信息。进一步，式(7)中描述的后验模型为不同移动场景下大规模MIMO***基站侧可获得的非完美CSI的一般模型。当αk非常接近1时，其适于用户准静止时的通信场景。当αk变得非常小时，其适于用户移动速度非常快的通信场景。进一步，此情形下
Figure PCTCN2017106351-appb-000020
变为接近零，并且式(7)中的后验模型和式(2)中的先验模型的差别变得非常小。根据用户不同的移动速度将αk设为不同的值，所建立后验模型可以用来描述大规模MIMO多种典型移动通信场景下的信道模型。
对于工作于FDD模式的大规模MIMO***，移动终端的信道估计、预测和反馈，同样可以获得式(7)中的后验联合相关模型。具体来说，基站发送下行全向导频；移动终端利用接收到的全向导频进行信道估计、预测与反馈。此时式(7)中所获得信道信息变为移动终端在接收到的下行导频信号条件下的条件均值(或称条件期望值)和条件方差信息。
四、鲁棒预编码设计
1、问题陈述
考虑时隙m上的下行传输。令xk,m,n表示时隙m第n块上第k个UE的Mk×1维发送向量，其协方差矩阵为单位阵。在时隙m第n块上一个符号间隔内，第k个UE的接收信号yk,m,n可以表示为
Figure PCTCN2017106351-appb-000021
其中Pk,m,n是第k个UE的Mk×dk维预编码矩阵，zk,m,n是一分布为
Figure PCTCN2017106351-appb-000022
的复高斯随机噪声向量,
Figure PCTCN2017106351-appb-000023
为噪声向量每一元素方差，
Figure PCTCN2017106351-appb-000024
为Mk×Mk单位矩阵。因为预编码矩阵Pk,m,n的设计基于后验统计模型，能够适应各种典型移动场景，即具有鲁棒性，所以将之称为鲁棒预编码。为降低***实现复杂度，只需在降维的鲁棒预编码域进行导频信号发送和数据信号发送。所发送的鲁棒预编码域导频信号在同一时频资源上，并且各用户导频不要求正交，即可以进行导频复用。具体而言，基站向各用户发送的预编码域导频信号为ZC序列或ZC序列组经过调制生成的频域信号。移动终端在接收到导频信号之后，进行鲁棒预编码域等效信道的信道估计，鲁棒预编码域等效信道为Hk,m,nPk,m,n。简单起见，假设UE端从可获得具有各自鲁棒预编码域等效信道矩阵的完美CSI。各用户在接收到数据信号后，利用接收到的数据信号可进行鲁棒预编码域信号检测。将每一UE的总干扰噪声
Figure PCTCN2017106351-appb-000025
视作高斯噪声。令Rk,m,n表示z'k,m,n的协方差矩阵，有
Figure PCTCN2017106351-appb-000026
其中期望函数
Figure PCTCN2017106351-appb-000027
{·}表示基于用户侧长时统计信息对Hk,m,n的期望函数。根据信道互易性，用户侧的长时统计信道信息和式(2)中给出的基站端长时统计信道信息一致。因此，期望函数
Figure PCTCN2017106351-appb-000028
{·}可以根据式(2)进行计算。假设第k个UE已知Rk,m,n，此时第k用户遍历速率可以表示为
Figure PCTCN2017106351-appb-000029
其中
Figure PCTCN2017106351-appb-000030
表示根据式(7)中后验模型得出的对于Hk,m,n的条件期望函数。定义函数
Figure PCTCN2017106351-appb-000031
表示加权遍历和速率，即为根据所建立后验统计信道模型计算出的加权和速率条件均值。本实施例的目的是设计预编码矩阵P1,m,n,P2,m,n,…,PK,m,n使其最大化加权遍历和速率，即求解优化问题
Figure PCTCN2017106351-appb-000032
其中wk是第k用户的加权因子，P为总功率约束。
2、用于鲁棒预编码设计的MM算法
优化问题(13)中目标函数是一个关于预编码矩阵极其复杂的函数，因此该问题很难直接求解。通过MM(Minorize Maximization or Majorize Minimization)算法可以将所述将加权遍历和速率最大化预编码设计问题转化为迭代求解二次型优化问题进行求解。MM算法的关键是找出目标函数的简单minorizing函数。简便起见，定义单边相关阵
Figure PCTCN2017106351-appb-000033
和
Figure PCTCN2017106351-appb-000034
为
Figure PCTCN2017106351-appb-000035
Figure PCTCN2017106351-appb-000036
其中
Figure PCTCN2017106351-appb-000037
且
Figure PCTCN2017106351-appb-000038
令
Figure PCTCN2017106351-appb-000039
定义为
Figure PCTCN2017106351-appb-000040
令Ek,m,n和
Figure PCTCN2017106351-appb-000041
分别为
Figure PCTCN2017106351-appb-000042
和
Figure PCTCN2017106351-appb-000043
接着，第k用户的速率均值Rk,m,n为
Figure PCTCN2017106351-appb-000044
定义函数
Figure PCTCN2017106351-appb-000045
为
Figure PCTCN2017106351-appb-000046
其中
Figure PCTCN2017106351-appb-000047
并且有
Figure PCTCN2017106351-appb-000048
Figure PCTCN2017106351-appb-000049
Figure PCTCN2017106351-appb-000050
Figure PCTCN2017106351-appb-000051
则g1是目标函数f在
Figure PCTCN2017106351-appb-000052
上的minorizing函数。利用g1，可以将原优化问题(13)转化为如下迭代问题
Figure PCTCN2017106351-appb-000053
式(26)中给出的预编码矩阵序列的极限点是原优化问题(13)的一个局部最大值点。进一步，式(26)中的优化问题为预编码矩阵P1,m,n,P2,m,n,…,PK,m,n的一个凹二次型函数。其最优解可以通过拉格朗日乘子法直接获得，为
Figure PCTCN2017106351-appb-000054
其中，μ*为能量约束对应的最优拉格朗日因子。观察式(27)以及式(22)至(25)，可得预编码的计算需要使用一些随机矩阵的期望。下文进一步给出如何利用确定性等同给出快速计算方法。
3、基于确定性等同的鲁棒预编码设计算算法
从式(23)和(24)中，可以观察出矩阵
Figure PCTCN2017106351-appb-000055
和
Figure PCTCN2017106351-appb-000056
与速率Rk,m,n关于
Figure PCTCN2017106351-appb-000057
和
Figure PCTCN2017106351-appb-000058
的导数密切相关。根据Rk,m,n的确定性等同可以推导
Figure PCTCN2017106351-appb-000059
和
Figure PCTCN2017106351-appb-000060
的确定性等同。定义
Figure PCTCN2017106351-appb-000061
Figure PCTCN2017106351-appb-000062
式(7)中给出的信道模型为一具有非零均值的联合相关模型。对于此类模型，可得Rk,m,n的 确定性等同为
Figure PCTCN2017106351-appb-000063
或者
Figure PCTCN2017106351-appb-000064
其中
Figure PCTCN2017106351-appb-000065
可以通过迭代方程获得
Figure PCTCN2017106351-appb-000066
Figure PCTCN2017106351-appb-000067
Figure PCTCN2017106351-appb-000068
Figure PCTCN2017106351-appb-000069
Figure PCTCN2017106351-appb-000070
Figure PCTCN2017106351-appb-000071
进一步可得，
Figure PCTCN2017106351-appb-000072
和
Figure PCTCN2017106351-appb-000073
的确定性等同
Figure PCTCN2017106351-appb-000074
Figure PCTCN2017106351-appb-000075
利用
Figure PCTCN2017106351-appb-000076
和
Figure PCTCN2017106351-appb-000077
的确定性等同进行快速计算，可得基于确定性等同的预编码设计为
Figure PCTCN2017106351-appb-000078
其中
Figure PCTCN2017106351-appb-000079
综上，可以将鲁棒预编码设计总结为如下步骤:
步骤1:设d＝0。随机生成一组预编码矩阵
Figure PCTCN2017106351-appb-000080
并将其进行归一化以满足总能量约束；
步骤2:根据式(16)计算按
Figure PCTCN2017106351-appb-000081
步骤3:根据式(34)-(35)计算Γk,m,n和
Figure PCTCN2017106351-appb-000082
步骤4:根据式(22),(38),(39)计算
Figure PCTCN2017106351-appb-000083
和
Figure PCTCN2017106351-appb-000084
步骤5:更新
Figure PCTCN2017106351-appb-000085
设d＝d+1；
重复步骤2到步骤5直到收敛或者达到一个预设目标。
五、实施效果
为了使本技术领域的人员更好地理解本发明方案，下面给出两种具体***配置下的本实施例中鲁棒预编码传输方法和已有方法遍历和速率性结果比较。
首先，给出本实施例中鲁棒预编码传输方法与BDMA方法的结果比较。考虑一配置为基站发送天线数量Mt＝128，用户数K＝10和用户天线数量Mk＝4的大规模MIMO***。用户的时间相关因子αk分为五种，依次为α1,α2＝0.999，α3,α4＝0.9，α5,α6＝0.5，α7,α8＝0.1和α9,α10＝0，代表了用户处于不同移动速度下的典型移动场景。图7给出了在所考虑大规模MIMO***下行链路下，本实施例中鲁棒预编码传输方法与BDMA方法的遍历和速率结果比较。从图7中可以看出，本实施例中鲁棒预编码传输方法明显优于BDMA方法。这是因为鲁棒预编码传输方法中基站利用包含信道均值(或称期望值)和方差信息的后验联合相关模型进行鲁棒预编码传输，而BDMA方法利用只包含信道方差信息的先验联合相关模型进行传输，未能充分利用基站可以获得的统计信道信息。
接着，给出本实施例中鲁棒预编码传输方法与鲁棒RZF预编码方法的结果比较。鲁棒RZF方法是单天线用户大规模MIMO***中广泛应用的RZF预编码方法在非完美信道状态信息下的扩展。考虑一配置为Mt＝128，K＝20和Mk＝1的大规模MIMO***。图8给出了在所考虑大规模MIMO***下行链路下，本实施例中鲁棒预编码传输方法与鲁棒RZF方法在三种不同移动场景下的遍历和速率结果比较。准静止场景下所有用户 αk＝0.999。移动较慢场景αk分为两种，一半为0.999，另一半为0.9。典型移动场景αk分为五种，依次为α1～α4＝0.999，α5～α8＝0.9，α9～α12＝0.5，α13～α16＝0.1和α17～α20＝0。从图8中，可以看出三种不同移动场景下本实施例中鲁棒预编码传输方法的性能都优于鲁棒RZF预编码方法。进一步，可以观察到性能增益在低SNR时较小，但是随着SNR增加逐渐变得显著。这表明和鲁棒RZF预编码方法相比，本实施例中鲁棒预编码传输方法能够更有效的抑制用户间干扰。
在本申请所提供的实施例中，应该理解到，所揭露的方法，在没有超过本申请的精神和范围内，可以通过其他的方式实现。当前的实施例只是一种示范性的例子，不应该作为限制，所给出的具体内容不应该限制本申请的目的。例如，一些特征可以忽略，或不执行。
本发明方案所公开的技术手段不仅限于上述实施方式所公开的技术手段，还包括由以上技术特征任意组合所组成的技术方案。应当指出，对于本技术领域的普通技术人员来说，在不脱离本发明原理的前提下，还可以做出若干改进和润饰，这些改进和润饰也视为本发明的保护范围。

Claims (20)

  1. 非完美信道状态信息下大规模MIMO鲁棒预编码传输方法，其特征在于，包括：基于导频信号和先验统计信道模型，基站或发送装置获得移动终端或接收装置信道的后验统计信道模型，包含信道均值或期望值、以及方差信息；基站或发送装置利用包含信道均值或期望值、以及方差信息的后验统计信道模型进行鲁棒预编码传输。
  2. 根据权利要求1所述的非完美信道状态信息下大规模MIMO鲁棒预编码传输方法，其特征在于，所述先验统计相关信道模型通过以下步骤获得：
    基站或发送装置通过上行信道探测获得；
    或
    通过移动终端或接收装置基于下行信道探测获得。
  3. 根据权利要求1所述的非完美信道状态信息下大规模MIMO鲁棒预编码传输方法，其特征在于，所述先验统计相关信道模型采用以下模型中的一种：联合相关信道模型，分离相关模型和全相关模型。
  4. 根据权利要求1所述的非完美信道状态信息下大规模MIMO鲁棒预编码传输方法，其特征在于，所述后验统计信道模型通过以下步骤获得：
    基站或发送装置利用上行导频信号和先验联合相关信道模型，通过信道估计和预测获得信道信息；
    或
    通过移动终端或接收装置利用下行导频信号和先验联合相关信道模型，基于信道估计、预测、反馈获得信道信息。
  5. 根据权利要求1所述的非完美信道状态信息下大规模MIMO鲁棒预编码传输方法，其特征在于，所述后验统计信道模型中信道均值或期望值、以及方差信息包括信道后验均值或后验期望值、以及后验方差信息。
  6. 根据权利要求5所述的非完美信道状态信息下大规模MIMO鲁棒预编码传输方法，其特征在于，所述信道后验均值或期望值、以及后验方差信息包括：
    基站或发送装置在接收到的上行导频信号条件下的条件均值或条件期望值、以及条件方差信息；
    或
    移动终端或接收装置在接收到的下行导频信号条件下的条件均值或条件期望值、以及条件方差信息。
  7. 根据权利要求1所述的非完美信道状态信息下大规模MIMO鲁棒预编码传输方法，其特征在于，所述后验统计信道模型为包含信道估计误差、信道老化和空间相关影响的后验统计信道模型。
  8. 根据权利要求1所述的非完美信道状态信息下大规模MIMO鲁棒预编码传输方法，其特征在于，所述后验统计相关信道模型采用以下模型中的一种：联合相关信道模型，分离相关模型和全相关模型。
  9. 根据权利要求1所述的非完美信道状态信息下大规模MIMO鲁棒预编码传输方法，其特征在于，在所述的鲁棒预编码传输中，基站或发送装置根据加权遍历和速率最大化准则，进行各移动终端或接收装置的线性预编码矩阵设计，加权遍历和速率为根据所建立后验统计信道模型计算出的加权和速率条件均值。
  10. 根据权利要求1所述的非完美信道状态信息下大规模MIMO鲁棒预编码传输方法，其特征在于，在所述的鲁棒预编码传输过程中，基站或发送装置根据加权遍历和速率最大化准则进行各移动终端或接收装置的线性预编码设计时，通过MM算法将所述将加权遍历和速率最大化预编码设计问题转化为迭代求解二次型优化问题进行求解。
  11. 根据权利要求10所述的非完美信道状态信息下大规模MIMO鲁棒预编码传输方法，其特征在于，所述二次型优化问题求解时所需的随机矩阵期望，利用其确定性等同进行快速计算。
  12. 非完美信道状态信息下大规模MIMO下行鲁棒预编码域导频复用信道信息获取方法，其特征在于，包括：基站或发送装置获得移动终端或接收装置信道的后验统计信道模型，包含信道均值或期望值、以及方差信息；基站或发送装置利用包含信道均值或期望值、以及方差信息的后验统计信道模型进行鲁棒预编码传输；在所述的鲁棒预编码传输中，下行链路在预编码域实施导频复用信道信息获取，基站或发送装置在预编码域向各移动终端或接收装置发送下行导频信号，移动终端或接收装置利用接收到的导频信号，进行预编码域等效信道的信道估计，预编码域等效信道为实际的传输信道乘以鲁棒预编码矩阵。
  13. 根据权利要求12所述的非完美信道状态信息下大规模MIMO下行鲁棒预编码域导频复用信道信息获取方法，其特征在于，基站或发送装置向各移动终端或接收装置发送的预编码域导频信号在同一时频资源上发送，各移动终端或接收装置的导 频不要求正交。
  14. 根据权利要求12所述的非完美信道状态信息下大规模MIMO下行鲁棒预编码域导频复用信道信息获取方法，其特征在于，基站或发送装置向各移动终端或接收装置发送的预编码域导频信号为ZC序列或ZC序列组经过调制生成的频域信号。
  15. 大规模MIMO鲁棒预编码传输的接收方法，其特征在于，包括：通过鲁棒预编码传输的发送信号经过传输信道后由移动终端或接收装置进行接收，移动终端或者接收装置利用接收到的发送信号进行接收信号处理。
  16. 根据权利要求15所述的大规模MIMO鲁棒预编码传输的接收方法，其特征在于，所述的接收到的发送信号包括下行全向导频信号，和/或鲁棒预编码域导频信号，和/或鲁棒预编码域数据信号。
  17. 根据权利要求16所述的大规模MIMO鲁棒预编码传输的接收方法，其特征在于，所述导频信号为ZC序列或ZC序列组经过调制生成的频域信号。
  18. 根据权利要求15所述的大规模MIMO鲁棒预编码传输的接收方法，其特征在于，在所述的接收信号处理中，移动终端或接收装置利用接收到的下行全向导频信号进行信道估计、预测和反馈。
  19. 根据权利要求15所述的大规模MIMO鲁棒预编码传输的接收方法，其特征在于，在所述的接收信号处理中，移动终端或接收装置利用接收到的鲁棒预编码域导频信号进行预编码域等效信道的信道估计。
  20. 根据权利要求15所述的大规模MIMO鲁棒预编码传输的接收方法，其特征在于，在所述的接收信号处理中，移动终端或接收装置利用接收到的鲁棒预编码域数据信号进行预编码域信号的解调或检测。
PCT/CN2017/106351 2017-08-31 2017-10-16 大规模mimo鲁棒预编码传输方法 WO2019041470A1 (zh)

Priority Applications (1)

Application Number Priority Date Filing Date Title
US16/643,560 US11177863B2 (en) 2017-08-31 2017-10-16 Massive MIMO robust precoding transmission method

Applications Claiming Priority (2)

Application Number Priority Date Filing Date Title
CN201710766951.3A CN107483088B (zh) 2017-08-31 2017-08-31 大规模mimo鲁棒预编码传输方法
CN201710766951.3 2017-08-31

Publications (1)

Publication Number Publication Date
WO2019041470A1 true WO2019041470A1 (zh) 2019-03-07

Family

ID=60604230

Family Applications (1)

Application Number Title Priority Date Filing Date
PCT/CN2017/106351 WO2019041470A1 (zh) 2017-08-31 2017-10-16 大规模mimo鲁棒预编码传输方法

Country Status (3)

Country Link
US (1) US11177863B2 (zh)
CN (1) CN107483088B (zh)
WO (1) WO2019041470A1 (zh)

Cited By (1)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN110635836A (zh) * 2019-09-12 2019-12-31 重庆大学 一种基于波束选择的毫米波大规模mimo***mmse-pca信道估计方法

Families Citing this family (16)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
US10707907B2 (en) * 2017-12-18 2020-07-07 Qualcomm Incorporated System and method for nonlinearity estimation with reference signals
CN108990167B (zh) * 2018-07-11 2021-09-07 东南大学 一种机器学习辅助的大规模mimo下行用户调度方法
CN110752866B (zh) * 2018-07-23 2021-08-20 华为技术有限公司 大规模多输入多输出mimo预编码传输方法以及装置
CN110061766A (zh) * 2019-04-18 2019-07-26 长春通视光电技术有限公司 基于鲁棒迭代再加权抑制噪声的信道估计方法
US11705943B1 (en) * 2019-08-30 2023-07-18 Intelligent Automation, Llc Distributed network control and link activation for multi-user MIMO communication
CN111193533B (zh) * 2019-12-05 2021-08-10 东南大学 大规模mimo波束域鲁棒预编码传输方法与***
CN113676225B (zh) * 2020-05-15 2022-10-04 华为技术有限公司 一种大规模mimo预编码传输方法与装置
CN114285446A (zh) 2020-09-27 2022-04-05 北京三星通信技术研究有限公司 大规模多输入多输出***中的传输方式切换方法及装置
CN112311706B (zh) * 2020-11-03 2021-11-16 南京大学 基于贝叶斯神经网络信道预测的频分双工下行传输方法
CN113839695B (zh) * 2021-09-16 2022-06-21 东南大学 Fdd大规模mimo和速率最优统计预编码方法及设备
CN113839696B (zh) * 2021-09-22 2022-07-19 东南大学 一种在线鲁棒的分布式多小区大规模mimo预编码方法
CN114143896A (zh) * 2021-12-10 2022-03-04 东南大学 一种大规模mimo跨频协作鲁棒传输方法
CN114978254B (zh) * 2022-05-12 2024-01-19 东南大学 基于ofdm***的机器学习辅助低复杂度鲁棒预编码算法
CN114900216A (zh) * 2022-05-18 2022-08-12 东南大学 大规模mimo鲁棒预编码器的迭代信干噪比设计方法
CN115459813B (zh) * 2022-07-29 2024-01-23 网络通信与安全紫金山实验室 Mimo***的性能优化方法、装置、设备及存储介质
CN117318771B (zh) * 2023-10-13 2024-07-02 重庆邮电大学 一种在非正交导频下实现mURLLC的***传输方法

Citations (4)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
US20080310523A1 (en) * 2007-06-14 2008-12-18 Telefonaktiebolaget Lm Ericsson (Publ) Method and Apparatus for Controlling Multi-Antenna Transmission in a Wireless Communication Network
CN104702543A (zh) * 2013-12-04 2015-06-10 华为技术有限公司 预编码方法及装置
CN105897627A (zh) * 2016-04-12 2016-08-24 东南大学 大规模mimo全向预编码传输下的迭代软干扰消除接收方法
CN106130697A (zh) * 2016-07-05 2016-11-16 重庆邮电大学 基于Bayes估计和数据流间功率分配的联合干扰相位对齐方法

Family Cites Families (7)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN101378277B (zh) * 2007-08-29 2016-04-06 上海贝尔股份有限公司 多用户预编码及调度方法和实现该方法的基站
US9100146B2 (en) * 2012-03-08 2015-08-04 Alcatel Lucent Virtual sectorization using an active anntenna array
CN104022977B (zh) * 2014-06-17 2017-03-08 东南大学 一种适用于mimo‑ofdm***的信道矩阵与干扰协方差矩阵估计方法
US10536249B2 (en) * 2015-10-15 2020-01-14 Ntt Docomo, Inc. Uplink pilot reuse and user-proximity detection in wireless networks
EP3353917B1 (en) * 2015-10-19 2020-02-05 Huawei Technologies Co., Ltd. Mitigating inter-cell pilot interference via network-based greedy sequence selection and exchange
US10447352B2 (en) * 2016-08-11 2019-10-15 National Instruments Corporation UE-aided channel reciprocity compensation for radio access in MIMO wireless communication systems
WO2019189558A1 (ja) * 2018-03-30 2019-10-03 三菱マテリアル株式会社 電子・電気機器用銅合金、電子・電気機器用銅合金板条材、電子・電気機器用部品、端子、及び、バスバー

Patent Citations (4)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
US20080310523A1 (en) * 2007-06-14 2008-12-18 Telefonaktiebolaget Lm Ericsson (Publ) Method and Apparatus for Controlling Multi-Antenna Transmission in a Wireless Communication Network
CN104702543A (zh) * 2013-12-04 2015-06-10 华为技术有限公司 预编码方法及装置
CN105897627A (zh) * 2016-04-12 2016-08-24 东南大学 大规模mimo全向预编码传输下的迭代软干扰消除接收方法
CN106130697A (zh) * 2016-07-05 2016-11-16 重庆邮电大学 基于Bayes估计和数据流间功率分配的联合干扰相位对齐方法

Cited By (1)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN110635836A (zh) * 2019-09-12 2019-12-31 重庆大学 一种基于波束选择的毫米波大规模mimo***mmse-pca信道估计方法

Also Published As

Publication number Publication date
CN107483088A (zh) 2017-12-15
CN107483088B (zh) 2021-05-04
US11177863B2 (en) 2021-11-16
US20200204226A1 (en) 2020-06-25

Similar Documents

Publication Publication Date Title
WO2019041470A1 (zh) 大规模mimo鲁棒预编码传输方法
TWI463823B (zh) 相關性衰落通道的模型基礎通道估計器和通道估計方法
CN108880774B (zh) 频分双工多用户大规模多天线***及其下行导频信号长度设计方法
CN107483091B (zh) 一种fdd大规模mimo-ofdm***下的信道信息反馈算法
US12021586B2 (en) Methods and apparatus for operating in a wideband communication system
WO2017219389A1 (zh) 大规模mimo***中实现完美全向预编码的同步信号和信号的发送与接收方法
EP2997705B1 (en) Large-scale fading coefficient estimation in wireless massive mimo systems
CN114642019B (zh) 一种信道信息获取的方法
CN104871437A (zh) Fdd***中信道互易性补偿方法和装置
US20100246715A1 (en) Wireless communication method and apparatus
US10756934B2 (en) Method for channel estimation in a wireless communication system, communication unit, terminal and communication system
KR101981087B1 (ko) 쌍방향 릴레이 장치 및 그에 의한 방법
CN110289898A (zh) 一种大规模mimo***中基于1比特压缩感知的信道反馈方法
EP2727304A2 (en) Noise estimation filter
CN104737481A (zh) 发送机和无线通信方法
WO2015112883A1 (en) System and method for early termination in iterative null-space directed singular value decomposition for mimo
EP3157216B1 (en) Method and device for acquiring downlink data in large-scale mimo system
CN103929274B (zh) 一种协作多点传输预编码方法
CN106921596B (zh) 一种用于ToA信息估计以及信道估计的方法、设备与***
Ferrand et al. Efficient channel state information acquisition in massive MIMO systems using non-orthogonal pilots
WO2020073342A1 (en) Non-linear precoding procedure
WO2013029561A1 (zh) 用于通信***的线性预编码的方法和装置
US10873405B2 (en) Method for obtaining uplink calibration values, calibration method, and corresponding terminal and base station
Choi et al. Achieving massive connectivity with non-orthogonal pilots and dirty-RFs
CN113330695B (zh) 信道估计中的数据符号的锚定过程

Legal Events

Date Code Title Description
121 Ep: the epo has been informed by wipo that ep was designated in this application

Ref document number: 17923112

Country of ref document: EP

Kind code of ref document: A1

NENP Non-entry into the national phase

Ref country code: DE

122 Ep: pct application non-entry in european phase

Ref document number: 17923112

Country of ref document: EP

Kind code of ref document: A1

122 Ep: pct application non-entry in european phase

Ref document number: 17923112

Country of ref document: EP

Kind code of ref document: A1

32PN Ep: public notification in the ep bulletin as address of the adressee cannot be established

Free format text: NOTING OF LOSS OF RIGHTS PURSUANT TO RULE 112(1) EPC (EPO FORM 1205A DATED 23.12.2020)

122 Ep: pct application non-entry in european phase

Ref document number: 17923112

Country of ref document: EP

Kind code of ref document: A1