WO2021109419A1 - 大规模mimo波束域鲁棒预编码传输方法与系统 - Google Patents

大规模mimo波束域鲁棒预编码传输方法与系统 Download PDF

Info

Publication number
WO2021109419A1
WO2021109419A1 PCT/CN2020/086168 CN2020086168W WO2021109419A1 WO 2021109419 A1 WO2021109419 A1 WO 2021109419A1 CN 2020086168 W CN2020086168 W CN 2020086168W WO 2021109419 A1 WO2021109419 A1 WO 2021109419A1
Authority
WO
WIPO (PCT)
Prior art keywords
domain
refined
refined beam
channel
precoding
Prior art date
Application number
PCT/CN2020/086168
Other languages
English (en)
French (fr)
Inventor
高西奇
卢安安
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 US17/629,181 priority Critical patent/US11646776B2/en
Publication of WO2021109419A1 publication Critical patent/WO2021109419A1/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
    • H04B7/06—Diversity systems; Multi-antenna system, i.e. transmission or reception using multiple antennas using two or more spaced independent antennas at the transmitting station
    • H04B7/0613—Diversity systems; Multi-antenna system, i.e. transmission or reception using multiple antennas using two or more spaced independent antennas at the transmitting station using simultaneous transmission
    • H04B7/0615—Diversity systems; Multi-antenna system, i.e. transmission or reception using multiple antennas using two or more spaced independent antennas at the transmitting station using simultaneous transmission of weighted versions of same signal
    • H04B7/0617—Diversity systems; Multi-antenna system, i.e. transmission or reception using multiple antennas using two or more spaced independent antennas at the transmitting station using simultaneous transmission of weighted versions of same signal for beam forming
    • 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
    • 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
    • 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
    • H04B7/06—Diversity systems; Multi-antenna system, i.e. transmission or reception using multiple antennas using two or more spaced independent antennas at the transmitting station
    • H04B7/0613—Diversity systems; Multi-antenna system, i.e. transmission or reception using multiple antennas using two or more spaced independent antennas at the transmitting station using simultaneous transmission
    • H04B7/0615—Diversity systems; Multi-antenna system, i.e. transmission or reception using multiple antennas using two or more spaced independent antennas at the transmitting station using simultaneous transmission of weighted versions of same signal
    • H04B7/0619—Diversity systems; Multi-antenna system, i.e. transmission or reception using multiple antennas using two or more spaced independent antennas at the transmitting station using simultaneous transmission of weighted versions of same signal using feedback from receiving side
    • H04B7/0621—Feedback content
    • H04B7/0634—Antenna weights or vector/matrix coefficients
    • 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
    • H04B7/06—Diversity systems; Multi-antenna system, i.e. transmission or reception using multiple antennas using two or more spaced independent antennas at the transmitting station
    • H04B7/0613—Diversity systems; Multi-antenna system, i.e. transmission or reception using multiple antennas using two or more spaced independent antennas at the transmitting station using simultaneous transmission
    • H04B7/0615—Diversity systems; Multi-antenna system, i.e. transmission or reception using multiple antennas using two or more spaced independent antennas at the transmitting station using simultaneous transmission of weighted versions of same signal
    • H04B7/0619—Diversity systems; Multi-antenna system, i.e. transmission or reception using multiple antennas using two or more spaced independent antennas at the transmitting station using simultaneous transmission of weighted versions of same signal using feedback from receiving side
    • H04B7/0636—Feedback format
    • H04B7/0639—Using selective indices, e.g. of a codebook, e.g. pre-distortion matrix index [PMI] or for beam selection

Definitions

  • the invention belongs to the field of communication technology, and relates to a method and system for robust precoding transmission in a large-scale MIMO beam domain.
  • MIMO Multiple-Input Multiple-Ouput
  • massive MIMO greatly improves the system capacity by equipping the base station (BS, BaseStation) with a large-scale antenna array, and makes full use of space dimensional resources.
  • B5G mobile communications
  • the establishment of the channel statistical model is the basis of the theoretical method of massive MIMO precoding transmission.
  • the limited antenna size limits the application of large-scale linear array antennas.
  • the base station is often equipped with large-scale area array antennas and other easier-to-implement antenna arrays, which in turn causes the number of antennas in a single dimension to be limited.
  • the traditional beam-domain channel model based on the Discrete Fourier Transform (DFT) matrix will deviate from the actual physical channel model to a considerable extent.
  • the base station in the massive MIMO wireless system is equipped with a large-scale antenna array, and the number of user antennas occupying the same time-frequency resources increases, which limits the time-frequency resources used for pilots.
  • DFT Discrete Fourier Transform
  • the present invention discloses a massive MIMO beam-domain robust precoding transmission method and system, which can solve the adaptability problem of the massive MIMO technology in various typical scenarios.
  • Massive MIMO beam-domain robust precoding transmission methods including:
  • the refined beam domain is a 1 times refined beam domain, an integer multiple or fractional multiple refined beam domain greater than 1, and the refined beam domain channels pass refined beam domains.
  • a posteriori statistical channel information containing refined beam-domain channel mean and variance information is used for robust precoding transmission.
  • the refined beam domain channel is multiplied by the user side refined sampling steering vector matrix to the left and right multiplied by the base station side refined sampling steering vector matrix conjugate matrix to obtain the antenna domain channel.
  • the refined beam domain prior statistical channel information is obtained by the base station through uplink channel detection; or, by the user terminal based on downlink channel detection.
  • the refined beam-domain posterior statistical channel information is obtained by the base station using uplink pilot signals and a priori refined beam-domain channel statistical information through channel estimation and prediction; or, the user terminal uses downlink pilot signals and A priori refined beam-domain statistical information is obtained based on channel estimation, prediction, and feedback.
  • the channel mean and variance information in the refined beam domain posterior statistical channel model is the refined beam domain channel posterior mean and posterior variance information; the channel posterior mean and posterior variance information includes:
  • the refined beam domain condition mean and condition variance information of the base station under the condition of the received uplink pilot signal are provided.
  • the refined beam-domain condition mean and conditional variance information of the user terminal under the condition of the received downlink pilot signal are provided.
  • the method for acquiring a priori statistical channel information in the refined beam domain includes: converting the pilot signal or channel information into a refined beam domain through a refined sampling steering vector matrix, and obtaining each user by using the refined beam domain sample statistics
  • the terminal refines the beam domain priori statistics on channel information.
  • using the refined beam domain sample statistics to obtain the refined beam domain prior statistical channel information of each user terminal is specifically: solving the channel energy according to the refined beam domain sample statistics and the equation of the channel energy matrix function matrix.
  • Matrix in the equation, only the channel energy matrix or the channel amplitude matrix are unknown matrices, and the remaining matrices are known matrices.
  • the base station performs linear precoding matrix design for each user terminal according to the weighted traversal and rate maximization criteria, and the weighted traversal and rate are based on the refined beam domain posterior statistical channel information The calculated weight and rate condition mean.
  • weighted traversal and rate maximization criterion is replaced with an upper bound of the weighted traversal and rate maximization criterion; or, the weighted traversal and the rate maximization criterion and the rate are replaced with their deterministic equivalent.
  • Robust precoding design methods for massive MIMO beam domain include:
  • the initial precoding is transferred to the refined beam domain through the refined sampling steering vector matrix;
  • the refined beam domain is a 1 times refined beam domain, an integer multiple or a fractional multiple refined beam domain greater than 1;
  • the refined beam domain precoding result is transferred to the antenna domain through the refined sampling steering vector matrix.
  • the method uses the truncated conjugate gradient method to design precoding, including:
  • the initial precoding is transferred to the refined beam domain through the refined sampling steering vector matrix
  • Steps (3)-(4) are repeated until the preset number of iterations or precoding convergence is reached, and the refined beam domain precoding is converted to the antenna domain precoding by using the refined sampling steering vector.
  • the sampled steering vector matrix is an oversampled DFT matrix when the steering vector is uniformly sampled and the number of samples is an integer multiple of the antenna.
  • a computing device including a memory, a processor, and a computer program stored on the memory and running on the processor, the computer program is loaded into the processor to realize the massive MIMO beam-domain robust precoding transmission Method, or the described robust precoding design method of massive MIMO beam domain based on truncated conjugate gradient method.
  • a massive MIMO beam-domain robust precoding transmission system includes a base station and multiple user terminals, and the base station is used for:
  • the refined beam domain is a 1 times refined beam domain, an integer multiple or fractional multiple refined beam domain greater than 1, and the refined beam domain channels pass refined beam domains.
  • a posteriori statistical channel information containing refined beam-domain channel mean and variance information is used for robust precoding transmission.
  • a massive MIMO beam-domain robust precoding transmission system includes a base station and multiple user terminals, and the base station is provided with the computing device.
  • the massive MIMO beam-domain downlink robust transmission method proposed by the present invention can solve the universality problem of massive MIMO to various typical mobile scenarios and achieve high spectrum efficiency.
  • a posteriori statistical channel information including refined beam-domain channel mean and variance information is used for robust precoding transmission.
  • the statistical channel information used is sparse and can be implemented with low complexity.
  • the robust precoding method can achieve dimensionality reduction transmission, which can reduce the pilot overhead required during data transmission, reduce the complexity of demodulation or detection, and improve the overall efficiency of transmission.
  • Figure 1 is a flowchart of a robust precoding transmission method for massive MIMO beam domain based on uplink detection
  • Figure 2 is a flow chart of a massive MIMO beam-domain robust precoding transmission method based on user feedback
  • Figure 3 is a flow chart of the truncated conjugate gradient method for robust precoding design in the massive MIMO beam domain
  • Figure 4 is a schematic diagram of the comparison of traversal and rate results between the beam-domain robust precoding transmission method and the existing method
  • Figure 5 is a schematic diagram of the comparison of the traversal and rate results of the beam-domain robust precoding transmission method under several different refinement multiples.
  • the massive MIMO beam-domain robust transmission method based on uplink detection disclosed in the embodiment of the present invention includes that the base station obtains the refined beam-domain prior statistical channel information of each user terminal through uplink channel detection; the base station is based on uplink channel detection. Frequency signal and refined beam-domain prior statistical channel information to obtain the refined beam-domain posterior statistical channel information of each user terminal, including the posterior mean (or expected value) and variance; the base station uses the refined beam-domain channel mean and variance information The posterior statistical channel information for robust precoding transmission.
  • the massive MIMO beam-domain robust transmission method based on user feedback disclosed in the embodiment of the present invention includes the user terminal obtaining respective refined beam-domain prior statistical channel information through downlink channel detection; the user terminal uses the downlink pilot Frequency signal and refined beam-domain prior statistical channel information, through channel estimation and prediction, the refined beam-domain posterior statistical channel information of each channel is obtained and fed back to the base station.
  • the refined beam-domain posterior statistical channel information includes the channel Mean and variance information; the base station uses the refined beam-domain posterior statistical channel information containing channel mean and variance information for robust precoding transmission.
  • the user terminal in the above embodiment can be a mobile terminal or a fixed terminal such as a mobile phone, a vehicle-mounted device, or a smart device; the refined beam domain channel is multiplied by the user-side refined sampling steering vector matrix on the left and the base station side refined sampling steering vector matrix is on the right After the conjugate matrix, the antenna domain channel can be obtained.
  • the refined beam domain is a 1 times refined beam domain, an integer multiple or a fractional multiple refined beam domain greater than 1, and the multiple refers to the ratio of the number of beams to the number of antennas.
  • the refined beam domain sample statistics can be used to obtain the refined beam domain prior statistical channel information of each user terminal.
  • the channel mean and variance information in the refined beam domain posterior statistical channel model is the refined beam domain channel posterior mean and posterior variance information, including the refined beam domain condition average and sum of the base station under the condition of the received uplink pilot signal Conditional variance information; or refined beam-domain condition mean and conditional variance information under the condition of the received downlink pilot signal by the user terminal.
  • a massive MIMO beam-domain robust precoding design method disclosed in an embodiment of the present invention uses a truncated conjugate gradient method to design precoding, including: (1) Initial precoding (either externally input or Generated by random method) Transfer to the refined beam domain through the refined sampling steering vector matrix; (2) Use the posterior statistical channel information to perform the initial conjugate gradient sparse calculation in the refined beam domain; (3) Perform in the refined beam domain Conjugate gradient update direction sparse calculation; (4) Perform refined beam-domain conjugate gradient calculation and update refined beam-domain precoding; repeat steps (3)-(4) until the preset number of iterations is reached or the precoding is converged and used
  • the refined sampling steering vector converts the refined beam domain precoding to antenna domain precoding.
  • 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 downlink robust precoding transmission method related to the refined beam domain of the present invention will be described in detail below with reference to specific communication system examples. It should be noted that the method of the present invention is not only applicable to the specific system models mentioned in the following examples, but also Applicable to system models of other configurations.
  • the massive MIMO system consists of a base station and K mobile terminals.
  • M k' M h'M v' .
  • the system time resource is divided into several time slots, each time slot includes N b time blocks, and each time block includes T symbol intervals.
  • the massive MIMO system considered in this embodiment works in a Time Division Duplexing (TDD) mode.
  • TDD Time Division Duplexing
  • the downlink transmission includes precoding domain pilot and data signal transmission.
  • the uplink pilot signal is only transmitted in the first time block.
  • the second to N b time blocks are used for downlink precoding domain pilot and data signal transmission.
  • the length of the uplink training sequence is the length of the block, that is, T symbol intervals.
  • 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.
  • the downlink omnidirectional pilot signal is transmitted in the first block, and feedback from the mobile terminal is received.
  • the steering vectors in U k are not required to be orthogonal to each other.
  • u i (i-1)/M k , and N k is a positive integer multiple greater than 1
  • U k is an oversampled DFT matrix.
  • G k,m,n (M k ⁇ W k,m,n ) is a refined beam domain channel matrix with independent elements, each row of which corresponds to the refined beam domain on the user side, and each column corresponds to the two-dimensional fine space on the base station side Reduced beam domain, M k is the refined beam domain channel amplitude matrix, W k, m, n is a random matrix composed of independent and identically distributed complex Gaussian random variables.
  • U k in equation (7) can be replaced by a unit array.
  • the method of the present invention is not only applicable to large-scale uniform area array antennas, but also applicable to other forms of antennas, such as cylindrical array antennas, area array antennas whose array elements are polarized antennas.
  • U k can be changed to the steering vector matrix of the corresponding array.
  • the refined beam-domain statistical model has more statistical characteristic directions, so it can more accurately characterize the actual physical channel model.
  • the obtained uplink channel statistical information can be directly used as the downlink channel statistical information.
  • the user side can obtain downlink statistical channel information and feed it back to the base station.
  • a method for obtaining priori statistical channel information in a refined beam domain is given below. Assuming that X k is the pilot matrix of the k-th user, the pilot matrix can be used to obtain prior statistical channel information.
  • the pilot matrix between users is orthogonal, and the pilots between different antennas do not need to be orthogonal, that is, X k does not need to be unitary matrix.
  • Y m,1 denote the pilot signal received by the base station on the first block of the m-th time slot, and
  • the superscript T represents transpose
  • the superscript * represents conjugate
  • the superscript H represents conjugate transpose
  • Z m,1 is a random matrix composed of independent and identically distributed complex Gaussian random variables. Since the pilot matrix of each user is orthogonal, the Left multiplication Multiply right Available
  • Equation (13) can be expressed as
  • T kr and T t are known matrices, and O kr NO t are also known matrices. Therefore, the only unknown parameter matrix on the right side of the equal sign of the above equation is the refined beam domain channel energy matrix ⁇ k . Therefore, the acquisition of the channel energy matrix ⁇ k is based on the sample statistical matrix ⁇ k and the determination matrix T kr , T t and O kr NO t . T kr ⁇ k T t +O kr NO t is called the function matrix of the channel energy matrix. Equation (16) belongs to the parameter matrix estimation problem.
  • c 0 is a constant independent of M k.
  • J is a matrix of all ones.
  • the first half of the derivation is slightly more complicated, as
  • the iterative formula can be constructed as follows
  • a refined sampling beam domain channel amplitude matrix can be obtained.
  • the steps for obtaining refined beam-domain statistics channel information can be summarized as follows:
  • Step 1 Receive the pilot signal X k sent by each mobile terminal
  • Step 2 Multiply the received pilot signal Y m,1 and each local user pilot signal X k respectively to obtain
  • Step 3 Convert the multiplied pilot signal to the refined beam domain
  • Step 4 Use the refined beam domain sample statistics Perform the refined beam domain prior statistical channel information acquisition for each mobile terminal.
  • step 4 uses the refined beam domain sample statistics ⁇ k to obtain the refined beam domain prior statistical channel information of each mobile terminal.
  • the method can be further refined as follows:
  • Step 2 Initialize M k ;
  • Step 3 Iterative calculation Among them, A k should be updated as follows with M k:
  • the instantaneous channel information can also be acquired first, and then the instantaneous channel information can be used to refine the beam-domain priori statistical channel information.
  • the following gives a method for obtaining refined beam-domain statistical channel information ⁇ k when the channel information is known. Multiply H k, m, 1 to the left Multiply right Available
  • T kr becomes The KL divergence function of ⁇ k and the channel energy matrix function matrix T kr ⁇ k T t is simplified as
  • Step 1 Obtain the channel matrix H k,m,1 ;
  • Step 2 Convert the channel matrix to the refined beam domain
  • Step 3 Use the refined beam domain sample statistics Perform the refined beam domain prior statistical channel information acquisition for each mobile terminal.
  • step 4 uses the refined beam domain sample statistics ⁇ k to obtain the refined beam domain prior statistical channel information of each mobile terminal.
  • the method can be further refined as follows:
  • Step 1 According to Calculation
  • Step 2 Initialize M k ;
  • Step 3 Iterative calculation Among them, A k should be updated as follows with M k:
  • the channel information obtained in the first time block on time slot m-1 is used for transmission in the m-th time slot.
  • a first-order Gauss Markov model is used to describe the time-correlation model.
  • the refined beam-domain channel on the n-th time block of the m-th time slot can be expressed as
  • ⁇ k,m (N b +n-1) is the correlation factor of channels G k,m,n and G k,m-1,1 , and this function is a time correlation factor related to the user's moving speed.
  • ⁇ k, m is the correlation factor of channels G k,m,n and G k,m-1,1 .
  • the model in equation (8) is used for channel prediction.
  • precoding is performed on the entire time slot m. For simplicity, without considering the channel estimation error, assuming that accurate channel information of the refined beam domain channel matrix G k,m-1,1 can be obtained, the posterior statistical information of the refined beam channel on time slot m can be obtained as
  • ⁇ k,m and the channel on the entire time slot m are related to the correlation factor ⁇ k,m of H k,m-1,1 .
  • a feasible approach is to take the root mean square of all correlation factors ⁇ k,m on the time slot.
  • the refined beam-domain posterior statistical model on time slot m can be obtained as
  • the channel posterior statistical model in equation (10) needs to be further studied based on the channel estimation error model, time correlation model and a priori statistical model.
  • H k,m-1,1 is expressed as Then the fine chemical posterior statistical model can be further expressed as
  • the posterior beam domain channel model can be written as
  • P k,m is the M k ⁇ d k -dimensional precoding matrix of the k-th UE
  • z k,m is a distribution as The complex Gaussian random noise vector, Is the variance of each element of the noise vector, Is the M k ⁇ M k identity matrix.
  • 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 sent by the base station to each user is a frequency domain signal generated by modulating a ZC sequence or a ZC sequence group.
  • the mobile terminal After receiving the pilot signal, the mobile terminal performs channel estimation of the equivalent channel in the robust precoding domain, and the equivalent channel in the robust precoding domain is H k,m P k,m .
  • the UE can obtain a perfect CSI with their respective robust precoding domain equivalent channel matrix.
  • the received data signal can be used to perform robust precoding domain signal detection.
  • expected function represents the expected function of H k,m based on long-term statistical information on the user side.
  • the long-term statistical channel information on the user side is consistent with the long-term statistical channel information on the base station side given in equation (43). Therefore, the expectation function It can be calculated according to equation (43). Assuming that the k-th UE knows R k,m , the traversal rate of the k-th user can be expressed as
  • function represents the weighted traversal and rate, that is, the weighted and rate condition average calculated according to the established refined beam domain posterior statistical channel model.
  • the purpose of this embodiment is to design the precoding matrix P 1,m ,P 2,m ,...,P K,m to maximize the weighted traversal and rate, that is, to solve the optimization problem
  • the method of solving the optimization problem includes gradient method, conjugate gradient method, Newton iterative method, and iterative method based on the iterative formula obtained by KKT condition.
  • the objective function can also be replaced with its upper bound or deterministically equivalent. In order to explain the solution method of the optimization problem more clearly, the following takes the upper bound of the objective function as an example to give an optimization method.
  • the rate of each user in the objective function in the optimization problem (48) can be replaced by its upper bound
  • the matrices A k,m , B k,m and D m are respectively
  • N k,m (n+1) -L k,m (n+1)+ ⁇ n N k,m (n) (66)
  • Equation (57) or its corresponding truncated conjugate gradient method is the antenna domain precoding update method. Such methods can also be performed in the refined beam domain to reduce complexity.
  • the truncated conjugate gradient method is a general method for solving optimization problems. It is not only suitable for simplified formula (57), but also can be used to directly solve optimization problems (48), and it can also be used to solve robust precoding under other optimization goals. design. Take the truncated conjugate gradient method of simplified formula (57) as an example to illustrate the realization of the refined beam domain truncated conjugate gradient method.
  • the truncated conjugate gradient method can be implemented in the refined beam domain to further reduce the algorithm complexity. To further illustrate the calculation process of the refined beam-domain truncated conjugate gradient method, detailed steps are given below.
  • the specific steps are as follows: First, calculate the refined beam-domain precoding energy matrix for
  • the performance of the robust precoding transmission method in this embodiment in three different mobile scenarios is better than that of the RZF and SLNR precoding methods. Further, it can be observed that the performance gain is smaller at low SNR, but gradually becomes significant as the SNR increases. This shows that compared with the RZF and SLNR precoding methods, the robust precoding transmission method in this embodiment can suppress inter-user interference more effectively.
  • twice the refinement rate is significantly higher than 1 times refinement, that is, the performance when DFT is used as the spatial feature direction. Furthermore, it can be observed that the 4 times refinement performance is relatively weaker than the 2 times refinement performance gain. This indicates that in this simulation scenario, the channel information provided by the horizontal and vertical 2x refinement is sufficiently accurate for the precoding performance.
  • the embodiment of the present invention also discloses a computing device, including a memory, a processor, and a computer program stored on the memory and running on the processor, which is implemented when the computer program is loaded to the processor.
  • a computing device including a memory, a processor, and a computer program stored on the memory and running on the processor, which is implemented when the computer program is loaded to the processor.
  • the device includes a processor, a communication bus, a memory, and a communication interface.
  • the processor may be a general-purpose central processing unit (CPU), a microprocessor, an application-specific integrated circuit (ASIC), or one or more integrated circuits for controlling the execution of the program of the present invention.
  • the communication bus may include a path for transferring information between the above-mentioned components.
  • the communication interface uses any device such as a transceiver to communicate with other devices or communication networks.
  • the memory can be a read-only memory (ROM) or other types of static storage devices that can store static information and instructions, random access memory (RAM) or other types of dynamic storage devices that can store information and instructions, or it can be an electronic device.
  • EEPROM Erasing programmable read-only memory
  • CD-ROM Compact Disc-read-only memory
  • disk storage media or other magnetic storage devices
  • EEPROM Erasing programmable read-only memory
  • the memory can exist independently and is connected to the processor through a bus.
  • the memory can also be integrated with the processor.
  • the memory is used to store application program codes for executing the solution of the present invention, and the processor controls the execution.
  • the processor is used to execute the application program code stored in the memory, so as to implement the information acquisition method provided in the foregoing embodiment.
  • the processor may include one or more CPUs, or may include multiple processors, and each of these processors may be a single-core processor or a multi-core processor.
  • the processor here may refer to one or more devices, circuits, and/or processing cores for processing data (for example, computer program instructions).
  • the embodiment of the present invention also discloses a massive MIMO beam-domain robust precoding transmission system, including a base station and multiple user terminals.
  • the base station is used to: Statistical channel information; based on pilot signals and refined beam-domain prior statistical channel information to obtain the refined beam-domain posterior statistical channel information of each user terminal, including the refined beam-domain posterior mean and variance; use the refined beam-domain
  • the posterior statistical channel information of the channel mean and variance information is transmitted by robust precoding.
  • the embodiment of the present invention also discloses a massive MIMO beam-domain robust precoding transmission system, including a base station and multiple user terminals, and the base station is provided with the aforementioned computing device.

Landscapes

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

Abstract

本发明公开了大规模MIMO波束域鲁棒预编码传输方法与系统。该方法基于基站侧和用户侧精细化采样导向矢量矩阵，并考虑了移动性引起的信道老化的影响，获得的信道状态信息为包含信道均值和方差信息的精细化波束域后验统计信道信息。本发明中，基站利用精细化波束域后验统计信道信息进行鲁棒预编码传输，采用的精细化波束域后验统计信道信息所依赖的信道模型是基于精细化采样空间角度所对应的采用导向矢量矩阵的信道模型，所用统计信道信息更加充分和准确，能够解决天线尺寸受限下大规模MIMO对各种典型移动场景的普适性问题，并取得高频谱效率，所提出的鲁棒预编码设计利用波束域信道的稀疏性和采样导向矢量矩阵的结构特征，能够大幅降低计算复杂度。

Description

大规模MIMO波束域鲁棒预编码传输方法与系统 技术领域
本发明属于通信技术领域，涉及大规模MIMO波束域鲁棒预编码传输方法与系统。
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更高质量、更高传输率、更高移动性、更高用户密度、更低时延、更低能耗等场景。近年来，为大幅提高无线频谱的频谱效率和功率效率，大规模多输入多输出(MIMO，Multiple-Input Multiple-Ouput)技术被广泛研究。目前，大规模MIMO已被确认为5G的关键技术之一。大规模MIMO通过在基站(BS，BaseStation)配备大规模天线阵列极大的提高系统容量，充分利用了空间维度资源。未来，大规模MIMO仍将是5G之后移动通信(B5G)的研究热点。
信道统计模型的建立是大规模MIMO预编码传输理论方法的基础。在实际大规模MIMO无线系统中，有限的天线尺寸限制了大规模线阵天线的应用，基站侧常配置大规模面阵天线等更易于实现的天线阵列，进而造成单一维度上的天线数量受限，在此限制下传统的基于l离散傅里叶变换(DFT，Discrete Fourier Transform)矩阵的波束域信道模型会在相当程度上偏离实际物理信道模型。另一方面，大规模MIMO无线系统中基站配置大规模天线阵列，且占据相同时频资源的用户天线数量增多，限制了用于导频的时频资源，在导频资源受限的情况下瞬时信道估计误差无法避免，同时在中高速移动通信场景还存在基站侧所获瞬时信道信息老化等因素，因此发展能描述各种典型移动通信场景的统计信道模型具有重要意义。文献中相关工作大都考虑大规模线阵天线，多采用传统基于DFT的波束域信道模型，且都没有考虑基于先验统计模型和瞬时信道信息的后验统计模型。
在大规模MIMO无线传输系统中，下行多用户预编码传输理论方法是对抗多用户干扰并实现频谱效率增益的关键，因而是大规模MIMO无线传输系统最为核心的问题之一。传统多用户MIMO系统中，预编码方法主要分为线性预编码和非线性预编码方法。非线性预编码方法虽然可取得最优性能，但是其极高的复杂度限制了其在大规模MIMO系统中使用，而简单的典型线性正则化迫零(RZF，Regularized Zero Forcing)预编码方法，对瞬时信道准确性要求较高，无法适用于导频资源受限以及中高速移动通信场景等信道信息非理想场景。为取得接近最优性能，需从可获得的信道信息出发，研究最优线性预编码传输方法。
发明内容
发明目的：针对现有技术的不足，本发明公开了大规模MIMO波束域鲁棒预编码传输方 法与系统，能够解决大规模MIMO技术在各种典型场景下的适应性问题。
技术方案：为了达到上述目的，本发明提供如下技术方案：
大规模MIMO波束域鲁棒预编码传输方法，包括：
获取各用户终端精细化波束域先验统计信道信息；所述精细化波束域为1倍精细化波束域、大于1的整数倍或分数倍精细化波束域，精细化波束域信道通过精细化采样导向矢量矩阵与天线域信道进行转换；
基于导频信号和精细化波束域先验统计信道信息获取各用户终端精细化波束域后验统计信道信息，包括精细化波束域后验均值和方差；
利用包含精细化波束域信道均值和方差信息的后验统计信道信息进行鲁棒预编码传输。
进一步地，所述精细化波束域信道左乘用户侧精细化采样导向矢量矩阵并右乘基站侧精细化采样导向矢量矩阵共轭矩阵后得到天线域信道。
进一步地，所述精细化波束域先验统计信道信息由基站通过上行信道探测获得；或者，通过用户终端基于下行信道探测获得。
进一步地，所述精细化波束域后验统计信道信息由基站利用上行导频信号和先验精细化波束域信道统计信息，通过信道估计和预测获得；或者，通过用户终端利用下行导频信号和先验精细化波束域统计信息，基于信道估计、预测、反馈获得。
进一步地，所述精细化波束域后验统计信道模型中信道均值和方差信息为精细化波束域信道后验均值和后验方差信息；所述信道后验均值和后验方差信息包括：
基站在接收到的上行导频信号条件下的精细化波束域条件均值和条件方差信息；或者，
用户终端在接收到的下行导频信号条件下的精细化波束域条件均值和条件方差信息。
进一步地，所述精细化波束域先验统计信道信息获取方法包括：将导频信号或信道信息通过精细化采样导向矢量矩阵转换到精细化波束域，利用精细化波束域样本统计量获取各用户终端精细化波束域先验统计信道信息。
进一步地，所述利用所述精细化波束域样本统计量获取各用户终端精细化波束域先验统计信道信息具体为：根据精细化波束域样本统计量和信道能量矩阵函数矩阵的方程求解信道能量矩阵；所述方程中只有信道能量矩阵或信道幅度矩阵为未知矩阵，其余矩阵为已知矩阵。
进一步地，在所述的鲁棒预编码传输中，基站根据加权遍历和速率最大化准则，进行各用户终端的线性预编码矩阵设计，加权遍历和速率为根据精细化波束域后验统计信道信息计算出的加权和速率条件均值。
进一步地，将所述加权遍历和速率最大化准则替换为加权遍历和速率最大化准则上界；或，将所述加权遍历和速率最大化准则中和速率替换为其确定性等同。
大规模MIMO波束域鲁棒预编码设计方法，包括：
将初始预编码通过精细化采样导向矢量矩阵转入精细化波束域；所述精细化波束域为1倍精细化波束域、大于1的整数倍或分数倍精细化波束域；
在精细化波束域利用后验统计信道信息进行波束域预编码更新；其中所述精细化波束域后验统计信道信息基于导频信号和精细化波束域先验统计信道信息获取；
将精细化波束域预编码结果通精细化采样导向矢量矩阵转入天线域。
进一步地，所述方法利用截断共轭梯度方法设计预编码，包括：
(1)将初始预编码通过精细化采样导向矢量矩阵转入精细化波束域；
(2)在精细化波束域利用后验统计信道信息进行初始共轭梯度稀疏计算；
(3)在精细化波束域进行共轭梯度更新方向稀疏计算；
(4)进行精细化波束域共轭梯度计算并更新精细化波束域预编码；
重复步骤(3)-(4)直到达到预设迭代次数或预编码收敛，利用精细化采样导向矢量将精细化波束域预编码转为天线域预编码。
进一步地，所述采样导向矢量矩阵在导向矢量均匀采样且采样数为天线整数倍时为过采样DFT矩阵。
一种计算设备，包括存储器、处理器及存储在存储器上并可在处理器上运行的计算机程序，所述计算机程序被加载至处理器时实现所述的大规模MIMO波束域鲁棒预编码传输方法，或者所述的基于截断共轭梯度法的大规模MIMO波束域鲁棒预编码设计方法。
大规模MIMO波束域鲁棒预编码传输系统，包括基站和多个用户终端，所述基站用于：
获取各用户终端精细化波束域先验统计信道信息；所述精细化波束域为1倍精细化波束域、大于1的整数倍或分数倍精细化波束域，精细化波束域信道通过精细化采样导向矢量矩阵与天线域信道进行转换；
基于导频信号和精细化波束域先验统计信道信息获取各用户终端精细化波束域后验统计信道信息，包括精细化波束域后验均值和方差；
利用包含精细化波束域信道均值和方差信息的后验统计信道信息进行鲁棒预编码传输。
大规模MIMO波束域鲁棒预编码传输系统，包括基站和多个用户终端，所述基站设有所述的计算设备。
有益效果：与现有技术相比，本发明提出的大规模MIMO波束域下行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。利用包括精细化波束域信道均值和方差信息的后验统计信道信息进行鲁棒预编码传输，所用统计信道信息具有稀疏性，能够进行低复杂度实现。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为基于上行探测的大规模MIMO波束域鲁棒预编码传输方法流程图；
图2为基于用户反馈的大规模MIMO波束域鲁棒预编码传输方法流程图；
图3为大规模MIMO波束域鲁棒预编码设计的截断共轭梯度方法流程图；
图4为波束域鲁棒预编码传输方法与现有方法的遍历和速率结果比较示意图；
图5为波束域鲁棒预编码传输方法在几种不同精细化倍数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基于上行探测的大规模MIMO波束域鲁棒传输方法，包括基站通过上行信道探测获取各用户终端精细化波束域先验统计信道信息；基站基于上行导频信号和精细化波束域先验统计信道信息获取各用户终端精细化波束域后验统计信道信息，包括后验均值(或称期望值)和方差；基站利用包含精细化波束域信道均值和方差信息的后验统计信道信息进行鲁棒预编码传输。
如图2所示，本发明实施例公开的基于用户反馈的大规模MIMO波束域鲁棒传输方法，包括用户终端通过下行信道探测获得各自精细化波束域先验统计信道信息；用户终端利用下行导频信号和精细化波束域先验统计信道信息，通过信道估计和预测获得各自信道的精细化波束域后验统计信道信息并反馈给基站，该精细化波束域后验先验统计信道信息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精细化波束域后验统计信道信息进行鲁棒预编码传输。
上述实施例中的用户终端可以是手机、车载设备、智能装备等移动终端或固定终端；精细化波束域信道左乘用户侧精细化采样导向矢量矩阵并右乘基站侧精细化采样导向矢量矩阵共轭矩阵后可得天线域信道，精细化波束域为1倍精细化波束域、大于1的整数倍或分数倍精细化波束域，倍数是指波束个数与天线个数的比值。可利用精细化波束域样本统计量获取各用户终端精细化波束域先验统计信道信息。精细化波束域后验统计信道模型中信道均值和方差信息为精细化波束域信道后验均值和后验方差信息，包括基站在接收到的上行导频信号条件下的精细化波束域条件均值和条件方差信息；或用户终端在接收到的下行导频信号条件下的精细化波束域条件均值和条件方差信息。
如图3所示，本发明实施例公开的一种大规模MIMO波束域鲁棒预编码设计方法，利用 截断共轭梯度方法设计预编码，包括：(1)将初始预编码(可由外部输入或者通过随机方式生成)通过精细化采样导向矢量矩阵转入精细化波束域；(2)在精细化波束域利用后验统计信道信息进行初始共轭梯度稀疏计算；(3)在精细化波束域进行共轭梯度更新方向稀疏计算；(4)进行精细化波束域共轭梯度计算并更新精细化波束域预编码；重复步骤(3)-(4)直到达到预设迭代次数或预编码收敛并利用精细化采样导向矢量将精细化波束域预编码转为天线域预编码。
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系统。下面结合具体的通信系统实例对本发明涉及精细化波束域下行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系统模型，也同样适用于其它配置的系统模型。
一、系统配置
考虑一平坦块衰落大规模MIMO系统，假设系统中各移动终端信道在T个符号间隔内保持不变。该大规模MIMO系统由一个基站和K个移动终端构成。基站配置大规模均匀面阵天线，每行和每列的天线数量分别为M h'和M v'，总发送天线数量为M t＝M h'M v'。简便起见，假设每个用户配置的天线数为M k'。将系统时间资源分为若干时隙，每一时隙包括N b个时间块，而每个时间块则包含T个符号间隔。本实施例中所考虑大规模MIMO系统工作于时分双工(TDD，Time Division Duplexing)模式。简便起见，假设只存在上行信道训练和下行传输阶段，下行传输包括预编码域导频和数据信号发送。在每一时隙中，只在第一时间块传输上行导频信号。第2至N b时间块则用于下行预编码域导频和数据信号传输。上行训练序列的长度为块的长度，即T个符号间隔。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、精细化波束域先验统计信道模型
下面对基于精细化采样导向矢量矩阵的精细化波束域先验统计模型进行详细阐述。定义基站侧水平和垂直维度上的精细化因子分别为N h和N v，令M h＝N hM h'和M v＝N vM v'分别表示水平和垂直维度上精细化采样导向矢量数量。传统波束域信道模型中，导向矢量和天线个数相同。对于所述精细化模型，采用更加精细的导向矢量，N h和N v为大于等于1的整数或分数。进一步，定义
Figure PCTCN2020086168-appb-000001
和
Figure PCTCN2020086168-appb-000002
则水平和垂直维度上的导向矢量矩阵可以分别表示为
Figure PCTCN2020086168-appb-000003
和
Figure PCTCN2020086168-appb-000004
其中导向矢量矩阵
Figure PCTCN2020086168-appb-000005
和
Figure PCTCN2020086168-appb-000006
中的导向矢量不要求互相正交。当导向矢量矩阵中矢量采用均匀采样，即v n＝(n-1)/M h和u m＝(m-1)/M v，N h和N v为大于1的正整数倍时，
Figure PCTCN2020086168-appb-000007
和
Figure PCTCN2020086168-appb-000008
为过采样DFT矩阵。令
Figure PCTCN2020086168-appb-000009
表示基站侧精细化采样导向矢量矩阵，则
Figure PCTCN2020086168-appb-000010
中的导向矢量也不要求互相正交。相似的，定义用户侧精细化因子为N k，其为大于或等于1的整数或分数。令M k＝N kM k'表示用户侧精细化采样导向矢量数量。进一步，定义
Figure PCTCN2020086168-appb-000011
则用户侧精细化采样导向矢量矩阵定义为
Figure PCTCN2020086168-appb-000012
同样，U k中的导向矢量也不要求互相正交。当导向矢量矩阵中矢量采用均匀采样，即u i＝(i-1)/M k，N k为大于1的正整数倍时，U k为过采样DFT矩阵。令H k,m,n表示第k用户在第m时隙第n块上的信道，则所考虑大规模MIMO系统波束域先验统计信道模型可以表示为
Figure PCTCN2020086168-appb-000013
其中G k,m,n＝(M k⊙W k,m,n)为元素独立的精细化波束域信道矩阵，其每一行对应用户侧精细化波束域，每一列对应基站侧二维空间精细化波束域，M k为精细化波束域信道幅度矩阵，W k,m,n为独立同分布复高斯随机变量组成的随机矩阵。当用户侧天线较少时，式(7)中U k可替换为单位阵。需要说明的是，本发明方法不仅适用于大规模均匀面阵天线，也适用于其他形式的天线，如圆柱阵列天线、阵元为极化天线的面阵天线，当基站侧或者用户侧使用的天线阵列发生变化时，
Figure PCTCN2020086168-appb-000014
或者U k改为相应阵列的导向矢量矩阵即可。与基于DFT矩阵的传统波束域先验统计信道模型相比，该精细化波束域统计模型有着更多的统计特征方向，因此能更准确地表征实际物理信道模型。定义大规模MIMO系统信道精细化波束域能量矩阵Ω k为Ω k＝M k⊙M k。
三、精细化波束域信道模型先验统计信道信息获取方法
对于所考虑工作于TDD模式的大规模MIMO系统，由于上下行信道具有互易性，获得的上行信道统计信息可以直接作为下行信道统计信息使用。对于FDD系统信道瞬时互易性不存在， 可以由用户侧进行下行统计信道信息获取并反馈给基站。下面给出一种精细化波束域先验统计信道信息获取的方法。假设X k为第k用户的导频矩阵，该导频矩阵可用于获取先验统计信道信息，用户间导频矩阵正交，不同天线间的导频不需要正交，即X k不必为酉矩阵。令Y m,1表示基站在第m时隙第1块上的接收到的导频信号，有
Figure PCTCN2020086168-appb-000015
进一步，有
Figure PCTCN2020086168-appb-000016
其中上标T表示转置，上标*表示共轭，上标H表示共轭转置，Z m,1为独立同分布复高斯随机变量组成的随机矩阵。由于各用户导频矩阵正交，将
Figure PCTCN2020086168-appb-000017
左乘
Figure PCTCN2020086168-appb-000018
并右乘
Figure PCTCN2020086168-appb-000019
可得
Figure PCTCN2020086168-appb-000020
其中⊙表示Hadmard乘积。令
Figure PCTCN2020086168-appb-000021
进一步，有
Figure PCTCN2020086168-appb-000022
令矩阵T kr表示
Figure PCTCN2020086168-appb-000023
矩阵T t表示
Figure PCTCN2020086168-appb-000024
矩阵O kr表示
Figure PCTCN2020086168-appb-000025
以及矩阵O t表示
Figure PCTCN2020086168-appb-000026
可以得到
Figure PCTCN2020086168-appb-000027
在噪声方差矩阵N已知情况下，则O krNO t为已知矩阵。简洁起见，令
Figure PCTCN2020086168-appb-000028
N＝N hN v为
Figure PCTCN2020086168-appb-000029
由于实际系统中只能获取样本平均，所以用重新定义Φ k为精细化波束域样本统计矩阵
Figure PCTCN2020086168-appb-000030
其中M表示样本数量。式(13)可按元素表示为
Figure PCTCN2020086168-appb-000031
利用Φ k，式(11)可以变为
Φ k＝T krΩ kT t+O krNO t         (16)
上式中，T kr和T t为已知矩阵，O krNO t同样为已知矩阵。因此，上述方程等号右边唯一的未知参数矩阵为精细化波束域信道能量矩阵Ω k。所以信道能量矩阵Ω k的获取基于样本统计矩阵Φ k和确定矩阵T kr、T t和O krNO t。将T krΩ kT t+O krNO t称做信道能量矩阵的函数矩阵。式(16)属于参数矩阵估计问题，为求解Ω k可根据式(16)建立优化问题，然后对优化问题利用梯度下降法、共轭梯度法、牛顿迭代法或者由KKT条件获得的迭代公式等方法进行求解。为了更加清晰说明该问题，下面给出一种具体估计方法。为估计出信道能量矩阵Ω k或者信道幅度矩阵M k，利用精细化波束域样本统计矩阵Φ k和信道能量矩阵函数矩阵T krΩ kT t+O krNO t的KL散度(divergence)定义目标函数为
Figure PCTCN2020086168-appb-000032
上式中c 0为和M k无关常数。为进行优化获得KL散度最小的M k，首先对目标函数求导，式(16)中后半部分的导数为
Figure PCTCN2020086168-appb-000033
其中J为全1矩阵。前半部分的求导稍微复杂，为
Figure PCTCN2020086168-appb-000034
其中，
Figure PCTCN2020086168-appb-000035
Figure PCTCN2020086168-appb-000036
综上，可得对g(M k)求导有，
Figure PCTCN2020086168-appb-000037
令g(M k)＝0，可得最优点必要条件为
(T tJT kr) T⊙M k-(T tQ TT kr) T⊙M k＝0         (22)
进一步有
(T tJT kr) T⊙M k＝(T tQ TT kr) T⊙M k      (23)
基于必要条件，可构造迭代公式如下
Figure PCTCN2020086168-appb-000038
其中，
Figure PCTCN2020086168-appb-000039
根据所提迭代公式可获得精细化采样波束域信道幅度矩阵。综上，精细化波束域统计信道信息获取的步骤可总结为：
步骤1：接收各移动终端发送的导频信号X k；
步骤2：将接收到的导频信号Y m,1与本地各用户导频信号X k分别相乘获得
Figure PCTCN2020086168-appb-000040
步骤3：将相乘后导频信号转换到精细化波束域
Figure PCTCN2020086168-appb-000041
步骤4：利用所述精细化波束域样本统计量
Figure PCTCN2020086168-appb-000042
进行各移动终端精细化波束域先验统计信道信息获取。
其中步骤4利用精细化波束域样本统计量Φ k进行各移动终端精细化波束域先验统计信道信息的获取方法可进一步细化为：
步骤1：计算
Figure PCTCN2020086168-appb-000043
及
Figure PCTCN2020086168-appb-000044
步骤2：初始化M k；
步骤3：迭代计算
Figure PCTCN2020086168-appb-000045
其中A k要随着M k做如下更新：
Figure PCTCN2020086168-appb-000046
前面讲述了利用导频信号进行精细化波束域先验统计信道信息获取的方法。在实际系统中也可先进行瞬时信道信息获取，然后利用瞬时信道信息进行精细化波束域先验统计信道信息。下面给出一种在信道信息已知情况下，精细化波束域统计信道信息Ω k获取的方法。将H k,m,1左乘
Figure PCTCN2020086168-appb-000047
并右乘
Figure PCTCN2020086168-appb-000048
可得
Figure PCTCN2020086168-appb-000049
进一步，有
Figure PCTCN2020086168-appb-000050
此时，精细化波束域样本统计矩阵
Figure PCTCN2020086168-appb-000051
变为
Figure PCTCN2020086168-appb-000052
或者按元素表示为
Figure PCTCN2020086168-appb-000053
进一步，可以得到
Φ k＝T krΩ kT t         (30)
此时，T kr变为
Figure PCTCN2020086168-appb-000054
Φ k和信道能量矩阵函数矩阵T krΩ kT t的KL散度函数简化为
Figure PCTCN2020086168-appb-000055
上式中c 0为和M k无关常数。同样，为进行优化获得KL散度最小的M k，首先对目标函数求导，式(31)中后半部分的导数变为
Figure PCTCN2020086168-appb-000056
其中J为全1矩阵。前半部分的求导变为
Figure PCTCN2020086168-appb-000057
其中，
Figure PCTCN2020086168-appb-000058
Figure PCTCN2020086168-appb-000059
综上，可得对g(M k)求导有，
Figure PCTCN2020086168-appb-000060
令g(M k)＝0，可得最优点必要条件为
(T tJT kr) TM k-(T tQ TT kr) T⊙M k＝0     (36)
进一步有
(T tJT kr) TM k＝(T tQ TT kr) T⊙M k     (37)
基于必要条件，构造迭代公式如下
Figure PCTCN2020086168-appb-000061
其中，
Figure PCTCN2020086168-appb-000062
综上，已知信道信息情形精细化波束域统计信道信息获取的步骤可总结为：
步骤1：获得信道矩阵H k,m,1；
步骤2：将信道矩阵转换到精细化波束域
Figure PCTCN2020086168-appb-000063
步骤3：利用所述精细化波束域样本统计量
Figure PCTCN2020086168-appb-000064
进行各移动终端精细化波束域先验统计信道信息获取。
其中步骤4利用精细化波束域样本统计量Φ k进行各移动终端精细化波束域先验统计信道信息的获取方法可进一步细化为：
步骤1：根据
Figure PCTCN2020086168-appb-000065
计算
Figure PCTCN2020086168-appb-000066
步骤2：初始化M k；
步骤3：迭代计算
Figure PCTCN2020086168-appb-000067
其中A k要随着M k做如下更新：
Figure PCTCN2020086168-appb-000068
三、精细化波束域后验统计信道模型
假设时隙m-1上第1时间块获得的信道信息用于第m时隙的传输。为描述大规模MIMO时间相关特性，采取一阶高斯马尔可夫模型来描述时间相关模型。在该模型下，第m时隙第n时间块上的精细化波束域信道可以表示为
Figure PCTCN2020086168-appb-000069
其中α k,m(N b+n-1)为信道G k,m,n和G k,m-1,1的相关因子，该函数和用户移动速度有关的时间相关因子。相关因子α k,m的获得方法有多种，这里假设相关因子已知。实际中，可以采用信道样本的经验相关因子，也可以采用文献中常用的基于Jakes自相关模型的相关因子α k,m的计算方法，即α k,m(n)＝J 0(2πv kf cnTτ/c)，其中J 0(·)表示第一类零阶Bessel函数，τ表示一个时间间隔对应的时间，v k表示第k用户速度，f c表示载波频率，c为光速。式(8)中模型用来进行信道预测。本实施例中，为考虑系统实现复杂度，在整个时隙m上进行预编码。简便起见，不考虑信道估计误差，假设可以获得精细化波束域信道矩阵G k,m-1,1的准确信道信息，可以得到时隙m上精细化波束信道的后验统计信息为
Figure PCTCN2020086168-appb-000070
其中β k,m和整个时隙m上信道与H k,m-1,1相关因子α k,m有关，一个可行的做法是取时隙上所有相关因子α k,m的均方根。进一步，则可以得到时隙m上的精细化波束域后验统计模型为
Figure PCTCN2020086168-appb-000071
当考虑信道估计误差时，式(10)中信道后验统计模型需要根据信道估计误差模型、时间相关模型和先验统计模型进一步进行研究得出。为便于在精细化波束域进行计算，将H k,m-1,1表示为
Figure PCTCN2020086168-appb-000072
则精细化后验统计模型可进一步表示为
Figure PCTCN2020086168-appb-000073
令
Figure PCTCN2020086168-appb-000074
进一步可将后验波束域信道模型写为
Figure PCTCN2020086168-appb-000075
其中
Figure PCTCN2020086168-appb-000076
为精细化波束域后验均值，(M' k,m⊙W k,m)的方差为精细化波束域后验方差。定义
Ω′ k,m＝M′ k,m⊙M′ k,m
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x k,m表示时隙m上第k个用户终端(UE)的M k×1维发送向量，其协方差矩阵为单位阵。在时隙m上，第k个UE的接收信号y k,m可以表示为
Figure PCTCN2020086168-appb-000077
其中P k,m是第k个UE的M k×d k维预编码矩阵，z k,m是一分布为
Figure PCTCN2020086168-appb-000078
的复高斯随机噪声向量,
Figure PCTCN2020086168-appb-000079
为噪声向量每一元素方差，
Figure PCTCN2020086168-appb-000080
为M k×M k单位矩阵。因为预编码矩阵P k,m的设计基于精细化波束域后验统计模型，能够适应各种典型大规模MIMO移动场景，即具有鲁棒性，所以将之称为精细化波束域下行鲁棒预编码。为降低系统实现复杂度，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H k,mP k,m。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 PCTCN2020086168-appb-000081
视作高斯噪声。令R k,m表示z' k,m的协方差矩阵，有
Figure PCTCN2020086168-appb-000082
其中期望函数
Figure PCTCN2020086168-appb-000083
表示基于用户侧长时统计信息对H k,m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(43)中给出的基站端长时统计信道信息一致。因此，期望函数
Figure PCTCN2020086168-appb-000084
可以根据式(43)进行计算。假设第k个UE已知R k,m，此时第k用户遍历速率可以表示为
Figure PCTCN2020086168-appb-000085
其中
Figure PCTCN2020086168-appb-000086
同样表示根据式(43)中后验模型得出的对于H k,m的条件期望函数。
定义函数
Figure PCTCN2020086168-appb-000087
表示加权遍历和速率，即为根据所建立的精细化波束域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P 1,m,P 2,m,…,P K,m使其最大化加权遍历和速率，即求解优化问题
Figure PCTCN2020086168-appb-000088
其中w k是第k用户的加权因子，P为总功率约束。该优化问题的求解方法有梯度法、共轭梯度法、牛顿迭代法以及根据KKT条件获得的迭代公式进行迭代的方法。此外，在具体实施中，目标函数还可以替换为其上界或者确定性等同。为了更加清晰的说明该优化问题的求解方法，下面以目标函数上界为例，给出一种优化方法。
2、低复杂鲁棒预编码设计方法
首先，优化问题(48)中目标函数中各用户的速率可以替换为其上界
Figure PCTCN2020086168-appb-000089
针对进行替换后的优化问题(48)，下面给出低复杂度鲁棒预编码设计方法。首先，定义
Figure PCTCN2020086168-appb-000090
为
Figure PCTCN2020086168-appb-000091
进一步，可以获得优化问题的KKT条件为
D mP k,m-A k,mP k,m+μP k,m＝0,k＝1,2,…,K       (51)
μ≥0         (52)
Figure PCTCN2020086168-appb-000092
其中，矩阵A k,m、B k,m和D m分别为
Figure PCTCN2020086168-appb-000093
Figure PCTCN2020086168-appb-000094
Figure PCTCN2020086168-appb-000095
通过KKT条件，可以得到鲁棒预编码的迭代公式为
Figure PCTCN2020086168-appb-000096
其中μ d的计算为
Figure PCTCN2020086168-appb-000097
为降低迭代算法中矩阵求逆
Figure PCTCN2020086168-appb-000098
引发的复杂度，采用截断共轭梯度法去求解，共轭梯度法步骤为：
首先，初始化P k,m(0)、L k,m(0)和N k,m(0)
Figure PCTCN2020086168-appb-000099
Figure PCTCN2020086168-appb-000100
N k,m(0)＝-L k,m(0)                (61)
接着，根据如下步骤进行迭代更新：
Figure PCTCN2020086168-appb-000101
P k,m(n+1)＝P k,m(n)+α nN k,m(n)       (63)
Figure PCTCN2020086168-appb-000102
Figure PCTCN2020086168-appb-000103
N k,m(n+1)＝-L k,m(n+1)+β nN k,m(n)        (66)
其中n表示第n迭代。式(57)或者其对应的截断共轭梯度法为天线域预编码更新方法。此类方法还可以在精细化波束域进行，以降低复杂度。截断共轭梯度方法为解决优化问题的一种通用方法，不仅适用于简化式(57)，也可以用来直接求解优化问题(48)，还可以用来求解其它优化目标下的鲁棒预编码设计。下面以简化式(57)的截断共轭梯度法为例，阐述精细化波束域 截断共轭梯度法实现。
3、精细化波束域截断共轭梯度方法详细步骤
截断共轭梯度法可在精细化波束域进行实现，以进一步降低算法复杂度。为进一步说明精细化波束域截断共轭梯度方法的计算过程，下面给出详细步骤。
步骤a):通过精细化采样导向矢量矩阵获取精细化波束域预编码。具体为：
Figure PCTCN2020086168-appb-000104
步骤b):利用精细化波束域后验稀疏统计信息进行初始共轭梯度稀疏计算。具体步骤阐述如下：首先，计算精细化波束域预编码能量矩阵
Figure PCTCN2020086168-appb-000105
为
Figure PCTCN2020086168-appb-000106
进一步，计算各用户的精细化波束域预编码能量矩阵之和
Figure PCTCN2020086168-appb-000107
为
Figure PCTCN2020086168-appb-000108
接着，计算精细化波束域第k用户波束域后验均值和第l用户波束域预编码矩阵之积
Figure PCTCN2020086168-appb-000109
为
Figure PCTCN2020086168-appb-000110
进一步，计算干扰加噪声协方差矩阵R k,m以及干扰噪声加信号协方差矩阵
Figure PCTCN2020086168-appb-000111
为
Figure PCTCN2020086168-appb-000112
Figure PCTCN2020086168-appb-000113
其中
Figure PCTCN2020086168-appb-000114
v为M k维全1列向量。接着，计算
Figure PCTCN2020086168-appb-000115
为
Figure PCTCN2020086168-appb-000116
其中
Figure PCTCN2020086168-appb-000117
的计算为
Figure PCTCN2020086168-appb-000118
此时，
Figure PCTCN2020086168-appb-000119
的计算公式变为
Figure PCTCN2020086168-appb-000120
令
Figure PCTCN2020086168-appb-000121
然后，计算Ω sum为
Figure PCTCN2020086168-appb-000122
以及
Figure PCTCN2020086168-appb-000123
为
Figure PCTCN2020086168-appb-000124
此时，
Figure PCTCN2020086168-appb-000125
的计算公式变为
Figure PCTCN2020086168-appb-000126
结合前面
Figure PCTCN2020086168-appb-000127
和
Figure PCTCN2020086168-appb-000128
的计算公式，进行精细化波束域初始共轭梯度稀疏计算。首先，计算
Figure PCTCN2020086168-appb-000129
为
Figure PCTCN2020086168-appb-000130
然后更新步长μ d
Figure PCTCN2020086168-appb-000131
最后，给出截断共轭梯度方法的精细化波束域域初始化矩阵为：
Figure PCTCN2020086168-appb-000132
Figure PCTCN2020086168-appb-000133
Figure PCTCN2020086168-appb-000134
步骤c):在精细化波束域进行共轭梯度更新方向稀疏计算。计算精细化波束域
Figure PCTCN2020086168-appb-000135
为
Figure PCTCN2020086168-appb-000136
步骤d)：进行精细化波束域共轭梯度计算并更新精细化波束域预编码。具体为按照如下步骤更新鲁棒预编码
Figure PCTCN2020086168-appb-000137
Figure PCTCN2020086168-appb-000138
Figure PCTCN2020086168-appb-000139
Figure PCTCN2020086168-appb-000140
Figure PCTCN2020086168-appb-000141
最后，重复c)、d)步骤直到达到预设迭代次数或预编码收敛，利用精细化采样导向矢量将精细化波束域预编码转为天线域预编码，即
Figure PCTCN2020086168-appb-000142
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，下面给出两种具体系统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RZF预编码方法的结果比较。RZF方法中假设采用的过时信道信息为准确信道信息。考虑一配置为M t＝128，K＝40和M k＝1的大规模MIMO系统，其中基站天线配置为M h'＝8,M v'＝16。简便起见，所有用户的移动速度设为相同，为30、120和250km/h。图4给出了在所考虑大规模MIMO系统下行链路下，本实施例中鲁棒预编码传输方法与RZF和SLNR方法在三种不同移动场景下的遍历和速率结果比较。从图9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RZF和SLNR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RZF和SLNR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。
接着，给出本实施例中鲁棒预编码传输方法在不同精细化倍数下的结果比较。保持所考虑大规模各项参数不变，用户移动速度设为120和250km/h。为体现不同精细化因子的影响，考虑3种精细化因子，分别为水平垂直1倍精细化、2倍精细化和4倍精细化。图5给出了在所考虑大规模MIMO系统下行链路下，本实施例中鲁棒预编码传输方法在三种精细化采样倍数下的遍历和速率结果比较。从图10中，可以看出两种移动场景下本实施例中鲁棒预编码传输方法的性能在三种精细化因子时都随着精细化倍数增加而提高。其中，两倍精细化速率显著高于1倍精细化即采用DFT作为空间特征方向时的性能。进一步，可以观察到4倍精细化性能较2倍精细化性能增益比较微弱。这表明此仿真场景下，水平垂直2倍精细化所提供的信道信息对于预编码性能而言已足够准确。
基于相同的发明构思，本发明实施例还公开了一种计算设备，包括存储器、处理器及存储在存储器上并可在处理器上运行的计算机程序，所述计算机程序被加载至处理器时实现上述的大规模MIMO波束域鲁棒预编码传输方法，或者大规模MIMO波束域鲁棒预编码设计方法。
在具体实现中，该设备包括处理器，通信总线，存储器以及通信接口。处理器可以是一个通用中央处理器(CPU)，微处理器，特定应用集成电路(ASIC)，或一个或多个用于控制本发明方案程序执行的集成电路。通信总线可包括一通路，在上述组件之间传送信息。通信接口，使用任何收发器一类的装置，用于与其他设备或通信网络通信。存储器可以是只读存储器(ROM)或可存储静态信息和指令的其他类型的静态存储设备，随机存取存储器(RAM)或者可存储信息和指令的其他类型的动态存储设备，也可以是电可擦可编程只读存储器(EEPROM)、只读光盘(CD-ROM)或其他光盘存储、盘存储介质或者其他磁存储设备、或者能够用于携带或存储具有指令或数据结构形式的期望的程序代码并能够由计算机存取的任何其他介质，但不限于此。存储器可以是独立存在，通过总线与处理器相连接。存储器也可以和处理器集成在一起。
其中，存储器用于存储执行本发明方案的应用程序代码，并由处理器来控制执行。处理器用于执行存储器中存储的应用程序代码，从而实现上述实施例提供的信息获取方法。处理器可以包括一个或多个CPU，也可以包括多个处理器，这些处理器中的每一个可以是一个单核处理器，也可以是一个多核处理器。这里的处理器可以指一个或多个设备、电路、和/或用于处理数据(例如计算机程序指令)的处理核。
基于相同的发明构思，本发明实施例还公开了一种大规模MIMO波束域鲁棒预编码传输系统，包括基站和多个用户终端，所述基站用于：获取各用户终端精细化波束域先验统计信道信息；基于导频信号和精细化波束域先验统计信道信息获取各用户终端精细化波束域后验统计信道信息，包括精细化波束域后验均值和方差；利用包含精细化波束域信道均值和方差信息的后验统计信道信息进行鲁棒预编码传输。
基于相同的发明构思，本发明实施例还公开了一种大规模MIMO波束域鲁棒预编码传输系统，包括基站和多个用户终端，所述基站设有上述的计算设备。
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 (15)

  1. 大规模MIMO波束域鲁棒预编码传输方法，其特征在于，包括：
    获取各用户终端精细化波束域先验统计信道信息；所述精细化波束域为1倍精细化波束域、大于1的整数倍或分数倍精细化波束域，精细化波束域信道通过精细化采样导向矢量矩阵与天线域信道进行转换；
    基于导频信号和精细化波束域先验统计信道信息获取各用户终端精细化波束域后验统计信道信息，包括精细化波束域后验均值和方差；
    利用包含精细化波束域信道均值和方差信息的后验统计信道信息进行鲁棒预编码传输。
  2. 根据权利要求1所述的大规模MIMO波束域鲁棒预编码传输方法，其特征在于，所述精细化波束域信道左乘用户侧精细化采样导向矢量矩阵并右乘基站侧精细化采样导向矢量矩阵共轭矩阵后得到天线域信道。
  3. 根据权利要求1所述的大规模MIMO波束域鲁棒预编码传输方法，其特征在于，所述精细化波束域先验统计信道信息由基站通过上行信道探测获得；或者，通过用户终端基于下行信道探测获得。
  4. 根据权利要求1所述的大规模MIMO波束域鲁棒预编码传输方法，其特征在于，所述精细化波束域后验统计信道信息由基站利用上行导频信号和先验精细化波束域信道统计信息，通过信道估计和预测获得；或者，通过用户终端利用下行导频信号和先验精细化波束域统计信息，基于信道估计、预测、反馈获得。
  5. 根据权利要求1所述的大规模MIMO波束域鲁棒预编码传输方法，其特征在于，所述精细化波束域后验统计信道模型中信道均值和方差信息为精细化波束域信道后验均值和后验方差信息；所述信道后验均值和后验方差信息包括：
    基站在接收到的上行导频信号条件下的精细化波束域条件均值和条件方差信息；或者，用户终端在接收到的下行导频信号条件下的精细化波束域条件均值和条件方差信息。
  6. 根据权利要求1所述的大规模MIMO波束域鲁棒预编码传输方法，其特征在于，所述精细化波束域先验统计信道信息获取方法包括：将导频信号或信道信息通过精细化采样导向矢量矩阵转换到精细化波束域，利用精细化波束域样本统计量获取各用户终端精细化波束域先验统计信道信息。
  7. 根据权利要求6所述的大规模MIMO波束域鲁棒预编码传输方法，其特征在于，所述利用所述精细化波束域样本统计量获取各用户终端精细化波束域先验统计信道信息具体为：根据精细化波束域样本统计量和信道能量矩阵函数矩阵的方程求解信道能量矩阵；所述方程中只有信道能量矩阵或信道幅度矩阵为未知矩阵，其余矩阵为已知矩阵。
  8. 根据权利要求1所述的大规模MIMO波束域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，基站根据加权遍历和速率最大化准则，进行各用户终端的线性预编码矩阵设计，加权遍历和速率为根据精细化波束域后验统计信道信息计算出的加权和速率条件均值。
  9. 根据权利要求8所述的大规模MIMO波束域鲁棒预编码传输方法，其特征在于，将所述加权遍历和速率最大化准则替换为加权遍历和速率最大化准则上界；或，将所述加权遍历和速率最大化准则中和速率替换为其确定性等同。
  10. 大规模MIMO波束域鲁棒预编码设计方法，其特征在于，包括：
    将初始预编码通过精细化采样导向矢量矩阵转入精细化波束域；所述精细化波束域为1倍精细化波束域、大于1的整数倍或分数倍精细化波束域；
    在精细化波束域利用后验统计信道信息进行波束域预编码更新；其中所述精细化波束域后验统计信道信息基于导频信号和精细化波束域先验统计信道信息获取；
    将精细化波束域预编码结果通精细化采样导向矢量矩阵转入天线域。
  11. 根据权利要求10所述的大规模MIMO波束域鲁棒预编码设计方法，其特征在于，所述方法利用截断共轭梯度方法设计预编码，包括：
    (1)将初始预编码通过精细化采样导向矢量矩阵转入精细化波束域；
    (2)在精细化波束域利用后验统计信道信息进行初始共轭梯度稀疏计算；
    (3)在精细化波束域进行共轭梯度更新方向稀疏计算；
    (4)进行精细化波束域共轭梯度计算并更新精细化波束域预编码；
    重复步骤(3)-(4)直到达到预设迭代次数或预编码收敛，利用精细化采样导向矢量将精细化波束域预编码转为天线域预编码。
  12. 根据权利要求10所述的大规模MIMO波束域鲁棒预编码设计方法，其特征在于，所述采样导向矢量矩阵在导向矢量均匀采样且采样数为天线整数倍时为过采样DFT矩阵。
  13. 一种计算设备，包括存储器、处理器及存储在存储器上并可在处理器上运行的计算机程序，其特征在于，所述计算机程序被加载至处理器时实现根据权利要求1-9任一项所述的大规模MIMO波束域鲁棒预编码传输方法，或者根据权利要求10-12任一项所述的基于截断共轭梯度法的大规模MIMO波束域鲁棒预编码设计方法。
  14. 大规模MIMO波束域鲁棒预编码传输系统，包括基站和多个用户终端，其特征在于，所述基站用于：
    获取各用户终端精细化波束域先验统计信道信息；所述精细化波束域为1倍精细化波束域、大于1的整数倍或分数倍精细化波束域，精细化波束域信道通过精细化采样导向矢量矩阵与天线域信道进行转换；
    基于导频信号和精细化波束域先验统计信道信息获取各用户终端精细化波束域后验统计信道信息，包括精细化波束域后验均值和方差；
    利用包含精细化波束域信道均值和方差信息的后验统计信道信息进行鲁棒预编码传输。
  15. 大规模MIMO波束域鲁棒预编码传输系统，包括基站和多个用户终端，其特征在于，所述基站设有根据权利要求13所述的计算设备。
PCT/CN2020/086168 2019-12-05 2020-04-22 大规模mimo波束域鲁棒预编码传输方法与系统 WO2021109419A1 (zh)

Priority Applications (1)

Application Number Priority Date Filing Date Title
US17/629,181 US11646776B2 (en) 2019-12-05 2020-04-22 Massive MIMO beam domain robust precoding transmission method and system

Applications Claiming Priority (2)

Application Number Priority Date Filing Date Title
CN201911232382.XA CN111193533B (zh) 2019-12-05 2019-12-05 大规模mimo波束域鲁棒预编码传输方法与系统
CN201911232382.X 2019-12-05

Publications (1)

Publication Number Publication Date
WO2021109419A1 true WO2021109419A1 (zh) 2021-06-10

Family

ID=70709412

Family Applications (1)

Application Number Title Priority Date Filing Date
PCT/CN2020/086168 WO2021109419A1 (zh) 2019-12-05 2020-04-22 大规模mimo波束域鲁棒预编码传输方法与系统

Country Status (3)

Country Link
US (1) US11646776B2 (zh)
CN (1) CN111193533B (zh)
WO (1) WO2021109419A1 (zh)

Cited By (1)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN117437973A (zh) * 2023-12-21 2024-01-23 齐鲁工业大学(山东省科学院) 一种单细胞转录组测序数据插补方法

Families Citing this family (6)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN112511201B (zh) * 2020-11-19 2021-10-26 东南大学 天波大规模mimo通信方法及模型和系统
CN113055061B (zh) * 2021-04-26 2022-07-15 中国科学技术大学 一种大规模mimo系统的低复杂度预编码方法
CN113890580B (zh) * 2021-09-17 2022-08-12 浙江大学 面向非对称毫米波大规模mimo的多用户上下行波束对准方法
CN114172549B (zh) * 2021-12-07 2022-06-24 东南大学 天波大规模mimo通信下行传输方法
CN114665932B (zh) * 2022-03-17 2023-03-10 东南大学 大规模mimo波束时延多普勒域统计信道信息获取方法
CN114826462B (zh) * 2022-04-22 2024-03-12 东南大学 面向大规模mimo统计端口选择的波束域信道增广方法

Citations (3)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
US20160134352A1 (en) * 2014-11-06 2016-05-12 Futurewei Technologies, Inc. System and Method for Beam-Formed Channel State Reference Signals
CN107483088A (zh) * 2017-08-31 2017-12-15 东南大学 大规模mimo鲁棒预编码传输方法
CN108683619A (zh) * 2018-05-18 2018-10-19 浙江大学城市学院 一种低复杂度的大规模mimo信道参数估计方法

Family Cites Families (11)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
US8204014B2 (en) * 2008-03-18 2012-06-19 Motorola Mobility, Inc. Method and system for codebook-based closed-loop MIMO using common pilots and analog feedback
CN102891815B (zh) * 2012-09-19 2015-06-17 北京航空航天大学 一种时分双工多基站协作系统中的低复杂度信道估计方法
KR20170032444A (ko) * 2014-07-31 2017-03-22 후아웨이 테크놀러지 컴퍼니 리미티드 데이터 전송 방법 및 장치
CN106375070B (zh) * 2015-07-24 2019-10-01 上海朗帛通信技术有限公司 一种用于移动通信系统中的重传方法和装置
EP3520242A4 (en) * 2016-09-30 2020-05-06 Nokia Technologies Oy DETERMINE THE MATCHING BETWEEN THE REFINING BEAM INDEX (BI) AND THE LOGICAL REFINING BEAM INDEX (BI) BY AN ACCESS NODE
US11405138B2 (en) * 2017-02-06 2022-08-02 Telefonaktiebolaget Lm Ericsson (Publ) Mixed space time and space frequency block coding
CN106899334B (zh) * 2017-04-13 2020-04-28 北京墨丘科技有限公司 一种通信方法及装置
US10637586B2 (en) * 2017-08-25 2020-04-28 National Science Foundation Turbo receivers for multiple-input multiple-output underwater acoustic communications
EP3751768A4 (en) * 2018-02-14 2021-01-20 Huawei Technologies Co., Ltd. COMMUNICATION PROCESS, COMMUNICATION DEVICE AND SYSTEM
CN109490894A (zh) * 2018-09-28 2019-03-19 天津大学 一种多波束精细化后处理方法
CN109981153B (zh) * 2019-04-11 2020-08-07 东南大学 一种人工噪声辅助的大规模mimo安全统计预编码方法

Patent Citations (3)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
US20160134352A1 (en) * 2014-11-06 2016-05-12 Futurewei Technologies, Inc. System and Method for Beam-Formed Channel State Reference Signals
CN107483088A (zh) * 2017-08-31 2017-12-15 东南大学 大规模mimo鲁棒预编码传输方法
CN108683619A (zh) * 2018-05-18 2018-10-19 浙江大学城市学院 一种低复杂度的大规模mimo信道参数估计方法

Non-Patent Citations (2)

* Cited by examiner, † Cited by third party
Title
LIU KAI; FENG HUI; YANG TAO; HU BO: "Robust Beamforming for Downlink 3D-MIMO Systems with l11-norm Bounded CSI Uncertainty", 2018 10TH INTERNATIONAL CONFERENCE ON WIRELESS COMMUNICATIONS AND SIGNAL PROCESSING (WCSP), IEEE, 18 October 2018 (2018-10-18), pages 1 - 5, XP033460261, DOI: 10.1109/WCSP.2018.8555867 *
LU AN-AN; GAO XIQI; ZHONG WEN; XIAO CHENGSHAN; MENG XIN: "Robust Transmission for Massive MIMO Downlink With Imperfect CSI", IEEE TRANSACTIONS ON COMMUNICATIONS, IEEE SERVICE CENTER, PISCATAWAY, NJ. USA., vol. 67, no. 8, 1 August 2019 (2019-08-01), PISCATAWAY, NJ. USA., pages 5362 - 5376, XP011739713, ISSN: 0090-6778, DOI: 10.1109/TCOMM.2019.2912383 *

Cited By (2)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN117437973A (zh) * 2023-12-21 2024-01-23 齐鲁工业大学(山东省科学院) 一种单细胞转录组测序数据插补方法
CN117437973B (zh) * 2023-12-21 2024-03-08 齐鲁工业大学(山东省科学院) 一种单细胞转录组测序数据插补方法

Also Published As

Publication number Publication date
US11646776B2 (en) 2023-05-09
CN111193533A (zh) 2020-05-22
US20220263554A1 (en) 2022-08-18
CN111193533B (zh) 2021-08-10

Similar Documents

Publication Publication Date Title
WO2021109419A1 (zh) 大规模mimo波束域鲁棒预编码传输方法与系统
CN111919394B (zh) 使用正交时频空分复用的无线系统中的预编码
Jiang et al. Achievable rates of FDD massive MIMO systems with spatial channel correlation
WO2021109420A1 (zh) 大规模mimo波束域统计信道信息获取方法与系统
US11177863B2 (en) Massive MIMO robust precoding transmission method
JP5393789B2 (ja) 適合コードブックを利用する方法、装置およびマシンアクセス可能媒体
EP2751947B1 (en) Spatially randomized pilot symbol transmission method, system and device for mimo wireless communications
US10511361B2 (en) Method for determining a precoding matrix and precoding module
Khan et al. A robust channel estimation scheme for 5G massive MIMO systems
WO2015112883A1 (en) System and method for early termination in iterative null-space directed singular value decomposition for mimo
Xu et al. Angular domain channel estimation for mmWave massive MIMO with one-bit ADCs/DACs
Jia et al. Massive beam-division multiple access for B5G cellular Internet of Things
Xiong et al. Performance analysis of uplink massive MIMO systems with variable‐resolution ADCs using MMSE and MRC detection
Wei et al. Joint channel parameter estimation in multi-cell massive MIMO system
WO2019157709A1 (zh) 信息获取方法、装置、设备和存储介质
CN111988069B (zh) 大规模mimo广义特征向量结构预编码求解方法、装置
CN110912588B (zh) 一种基于改进Prony方法的下行时变信道预测方法
CN113839695B (zh) Fdd大规模mimo和速率最优统计预编码方法及设备
CN114629533B (zh) 大规模mimo信道估计的信息几何方法及系统
CN113676225A (zh) 一种大规模mimo预编码传输方法与装置
Dai et al. A scalable generator for massive MIMO baseband processing systems with beamspace channel estimation
Kumeleh et al. Pilot contamination reduction using time-domain channel sparsity in massive MIMO-OFDM systems
CN114665932B (zh) 大规模mimo波束时延多普勒域统计信道信息获取方法
Kafedziski Compressed sensing based channel estimation in FDD multi-user massive MIMO using angle domain sparsity and transmit antenna correlation
Mateur et al. Sparsity adaptive subspace pursuit based channel estimation algorithm for OFDM based massive MIMO systems

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: 20896438

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: 20896438

Country of ref document: EP

Kind code of ref document: A1

122 Ep: pct application non-entry in european phase

Ref document number: 20896438

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 06/07/2023)

122 Ep: pct application non-entry in european phase

Ref document number: 20896438

Country of ref document: EP

Kind code of ref document: A1