CN102594489B

Movatterモバイル変換

Info

Publication number: CN102594489B
Application number: CN201110008647.5A
Authority: CN
Inventors: 潘成康; 王启星; 刘建军; 刘光毅
Original assignee: China Mobile Communications Group Co Ltd
Current assignee: China Mobile Communications Group Co Ltd
Priority date: 2011-01-14
Filing date: 2011-01-14
Publication date: 2014-12-10
Anticipated expiration: 2031-01-14
Also published as: CN102594489A

Abstract

本发明公开了一种预编码向量的确定方法及装置、基站及用户终端，以解决现有预编码向量确定方式消耗较多运算资源的问题。方法包括：确定各个用户终端的等效信道矩阵T_j，并针对每一用户终端执行：根据其他用户终端的T_j，确定干扰信道矩阵并对中的所有行向量执行正交化处理得到一组标准正交基B_i；根据B_i确定该用户终端的等效信道矩阵到所张成的子空间的正交投影矩阵确定该用户终端接收天线的数量，以作为待SVD的矩阵，从预先设置的接收天线数量与简化SVD算法的对应关系中，确定与确定的该数量对应的简化SVD算法；根据基于确定的简化SVD算法所确定的的奇异值对应的右奇异向量，确定针对该用户终端的预编码向量。

The invention discloses a method and device for determining a precoding vector, a base station and a user terminal, so as to solve the problem that the existing way of determining a precoding vector consumes more computing resources. The method includes: determining the equivalent channel matrix T_j of each user terminal, and executing for each user terminal: determining the interference channel matrix according to T_j of other user terminals and to Orthogonalization processing is performed on all row vectors in to obtain a set of orthonormal bases B_i ; according to B_i determine the equivalent channel matrix of the user terminal to orthographic projection matrix of the spanned subspace Determine the number of receiving antennas for the user terminal, to As the matrix to be SVD, from the preset corresponding relationship between the number of receiving antennas and the simplified SVD algorithm, determine the simplified SVD algorithm corresponding to the determined number; according to the determined simplified SVD algorithm based on the determined The right singular vector corresponding to the singular value of is used to determine the precoding vector for the user terminal.

Description

Translated fromChinese

预编码向量的确定方法、装置、基站与用户终端Method, device, base station and user terminal for determining precoding vector

技术领域technical field

本发明涉及无线通信技术领域，尤其涉及一种预编码向量的确定方法及装置、基站及用户终端。The present invention relates to the technical field of wireless communication, in particular to a method and device for determining a precoding vector, a base station and a user terminal.

背景技术Background technique

基于预编码的下行传输方案是多天线系统的核心传输方案之一。所谓预编码是基站将待传信息符号乘以一个预编码向量，该预编码向量中的元素与发送天线是一一对应的关系。预编码向量通常是基于奇异值分解(SVD，SingularValue Decomposition)的方式获得的，即对下行多入多出(MIMO，Multiple-Input Multiple-Out-put)信道矩阵H进行SVD分解，由H的前n个最大奇异值对应的右奇异向量直接得到预编码向量，这里n为同时传输的信息符号数。The downlink transmission scheme based on precoding is one of the core transmission schemes of the multi-antenna system. The so-called precoding means that the base station multiplies the information symbols to be transmitted by a precoding vector, and the elements in the precoding vector have a one-to-one correspondence relationship with the transmitting antennas. The precoding vector is usually obtained based on Singular Value Decomposition (SVD, Singular Value Decomposition), that is, the SVD decomposition is performed on the downlink Multiple-Input Multiple-Output (MIMO, Multiple-Input Multiple-Out-put) channel matrix H. The right singular vectors corresponding to the n largest singular values directly obtain the precoding vector, where n is the number of information symbols transmitted simultaneously.

在时分双工(TDD，Time Division Duplexing)系统中，基站利用用户终端发送的探测导频信号(SRS)和信道互易性获知基站到该用户终端的下行信道矩阵H，从而由基站直接对H进行SVD分解。而在频分双工(FDD，FrequencyDivision Duplexing)系统中，用户终端通过基站发送的下行公共导频信号获知下行信道矩阵H，从而由用户终端执行对H的SVD，并由用户终端反馈H的右奇异向量所对应的预编码码字索引(PMI)到基站，以便使基站能够确定相应于该PMI的预编码码字从而执行后续的预编码。In a time division duplex (TDD, Time Division Duplexing) system, the base station uses the sounding pilot signal (SRS) sent by the user terminal and the channel reciprocity to know the downlink channel matrix H from the base station to the user terminal, so that the base station directly compares H Perform SVD decomposition. In the frequency division duplexing (FDD, Frequency Division Duplexing) system, the user terminal obtains the downlink channel matrix H through the downlink common pilot signal sent by the base station, so that the user terminal performs SVD on H, and the user terminal feeds back the right matrix of H The precoding codeword index (PMI) corresponding to the singular vector is sent to the base station, so that the base station can determine the precoding codeword corresponding to the PMI to perform subsequent precoding.

在单用户终端传输情况下，基站执行单用户预编码；在多用户终端传输情况下，基站执行多用户预编码。如果是多用户预编码，则需要消除用户间的干扰，即需要进行迫零处理。迫零处理通常是采用求解零空间的方法来实现的，即首先求取干扰信道(其它同时传输的用户的信道统称干扰信道)矩阵的零空间(零空间又称正交补空间，由其一组标准正交基表征，可以采用求解线性方程组的方法获得标准正交基，干扰信道矩阵与其零空间正交)，然后再对用户信道矩阵在该零空间中的投影进行SVD分解，并通过这次SVD分解获得的右奇异向量来构造用户最终的预编码向量。In the case of single-user terminal transmission, the base station performs single-user precoding; in the case of multi-user terminal transmission, the base station performs multi-user precoding. In the case of multi-user precoding, interference between users needs to be eliminated, that is, zero-forcing processing needs to be performed. The zero-forcing process is usually realized by solving the null space, that is, firstly, the null space of the interference channel (the channels of other users transmitting at the same time are collectively referred to as the interference channel) matrix (the null space is also called the orthogonal complement space, by one group orthonormal basis representation, the orthonormal basis can be obtained by solving linear equations, the interference channel matrix is orthogonal to its null space), and then the SVD decomposition is performed on the projection of the user channel matrix in the null space, and through The right singular vector obtained by this SVD decomposition is used to construct the user's final precoding vector.

预编码技术在增强MIMO和协作多点(CoMP)系统中的应用在带来性能提升的同时，也带来了工程实现难的困境。这是因为在发送端天线数较多的情况下(例如，增强MIMO中为8根发送天线，CoMP系统中多达24根发送天线)，SVD分解和干扰信道矩阵零空间的求解运算量巨大，因此需要消耗非常多的运算资源。The application of precoding technology in enhanced MIMO and Coordinated Multipoint (CoMP) systems not only improves performance, but also brings difficulties in engineering implementation. This is because in the case of a large number of antennas at the transmitting end (for example, 8 transmitting antennas in enhanced MIMO, up to 24 transmitting antennas in a CoMP system), the calculation of SVD decomposition and the solution of the null space of the interference channel matrix is huge, Therefore, it needs to consume a lot of computing resources.

发明内容Contents of the invention

本发明实施例提供一种预编码向量的确定方法及装置，用以解决现有技术提供的确定预编码向量的方式会消耗较多运算资源的问题。Embodiments of the present invention provide a method and device for determining a precoding vector, which are used to solve the problem that the way of determining the precoding vector provided by the prior art consumes more computing resources.

本发明实施例还提供一种基站和一种用户终端。The embodiment of the present invention also provides a base station and a user terminal.

本发明实施例采用以下技术方案：Embodiments of the present invention adopt the following technical solutions:

一种预编码向量的确定方法，包括：A method for determining a precoding vector, comprising:

分别确定各个用户终端对应的等效信道矩阵T_j，并针对每一所述用户终端，执行下述步骤：Determine the equivalent channel matrix T_j corresponding to each user terminal, and perform the following steps for each user terminal:

根据除该用户终端外的其他用户终端所对应的T_j，确定该用户终端的干扰信道矩阵并对中的所有行向量执行正交化处理，得到所张成的子空间的一组标准正交基B_i；以及根据B_i确定该用户终端的等效信道矩阵到所张成的子空间的正交投影矩阵以及确定该用户终端接收天线的数量，并以作为待进行奇异值分解SVD的矩阵，从预先设置的接收天线数量与针对待进行SVD的矩阵执行简化SVD算法的对应关系中，确定与确定的该数量对应的简化SVD算法；根据基于确定的简化SVD算法所确定的的奇异值对应的右奇异向量，确定针对该用户终端的预编码向量。According to T_j corresponding to other user terminals except the user terminal, determine the interference channel matrix of the user terminal and to Orthogonalization is performed on all the row vectors in to get A set of orthonormal bases B_i of the subspace spanned; and determine the equivalent channel matrix of the user terminal according to B_i to orthographic projection matrix of the spanned subspace and determine the number of receiving antennas of the user terminal, and use As the matrix to be subjected to singular value decomposition SVD, from the corresponding relationship between the preset number of receiving antennas and the simplified SVD algorithm for the matrix to be performed on the SVD, determine the simplified SVD algorithm corresponding to the determined number; according to the simplified SVD algorithm based on the determination determined by the SVD algorithm The right singular vector corresponding to the singular value of is used to determine the precoding vector for the user terminal.

一种预编码向量的确定方法，应用于TDD系统的单用户终端传输方案中，包括：A method for determining a precoding vector, applied to a single-user terminal transmission scheme of a TDD system, comprising:

基站确定所述用户终端对应的下行信道矩阵H_i，并确定所述用户终端接收天线的数量；以及以H_i作为待进行SVD的矩阵，从预先设置的接收天线数量与针对待进行SVD的矩阵执行简化SVD算法的对应关系中，确定与确定的该数量对应的简化SVD算法；并根据确定的简化SVD算法确定H_i的奇异值对应的右奇异向量，确定针对所述用户终端的预编码向量。The base station determines the downlink channel matrix H_i corresponding to the user terminal, and determines the number of receiving antennas of the user terminal; and uses H_i as the matrix to be performed SVD, from the preset number of receiving antennas and the matrix for the SVD to be performed In the corresponding relationship of executing the simplified SVD algorithm, determine the simplified SVD algorithm corresponding to the determined number; and determine the right singular vector corresponding to the singular value of_Hi according to the determined simplified SVD algorithm, and determine the precoding vector for the user terminal .

一种预编码向量的确定方法，应用于FDD系统的单用户终端传输方案中，包括：A method for determining a precoding vector, applied to a single-user terminal transmission scheme of an FDD system, comprising:

所述用户终端确定自身的下行信道矩阵H_i，并确定所述用户终端接收天线的数量；以及以H_i作为待进行SVD的矩阵，从预先设置的接收天线数量与针对待进行SVD的矩阵执行简化SVD算法的对应关系中，确定与确定的该数量对应的简化SVD算法；并根据确定的简化SVD算法确定H_i的奇异值对应的右奇异向量，根据所述右奇异向量确定并向基站发送相应的PMI，指示基站根据PMI确定相应的预编码向量。The user terminal determines its own downlink channel matrix H_i , and determines the number of receiving antennas of the user terminal; and uses H_i as the matrix to perform SVD, and executes from the preset number of receiving antennas and the matrix for the SVD to be performed In the corresponding relationship of the simplified SVD algorithm, determine the simplified SVD algorithm corresponding to the determined number; and determine the right singular vector corresponding to the singular value of_Hi according to the determined simplified SVD algorithm, determine and send to the base station according to the right singular vector The corresponding PMI indicates that the base station determines the corresponding precoding vector according to the PMI.

一种预编码向量的确定装置，包括：等效信道矩阵确定单元，用于确定各个用户终端对应的等效信道矩阵T_j；干扰信道矩阵确定单元，用于针对每一用户终端，根据等效信道矩阵确定单元确定的除该用户终端外的其他用户终端所对应的等效信道矩阵T_j，确定该用户终端的干扰信道矩阵正交化处理单元，用于对干扰信道矩阵确定单元确定的中的所有行向量执行正交化处理，得到所张成的子空间的一组标准正交基B_i；正交投影矩阵确定单元，用于根据正交化处理单元得到的B_i，确定该用户终端的等效信道矩阵到所张成的子空间的正交投影矩阵简化SVD算法确定单元，用于根据该用户终端接收天线的数量，以正交投影矩阵确定单元确定的作为待进行奇异值分解SVD的矩阵，从预先设置的接收天线数量与针对待进行SVD的矩阵执行简化SVD算法的对应关系中，确定与该用户终端接收天线的数量对应的简化SVD算法；预编码向量确定单元，用于根据基于简化SVD算法确定单元确定的简化SVD算法所确定的的奇异值对应的右奇异向量，确定针对该用户终端的预编码向量。A device for determining a precoding vector, comprising: an equivalent channel matrix determining unit, configured to determine an equivalent channel matrix T_j corresponding to each user terminal; an interference channel matrix determining unit, configured for each user terminal, according to the equivalent The channel matrix determination unit determines the equivalent channel matrix T_j corresponding to other user terminals except the user terminal, and determines the interference channel matrix of the user terminal Orthogonalization processing unit, used to determine the interference channel matrix determination unit Orthogonalization is performed on all the row vectors in to get A set of orthonormal bases B_i of the subspace formed; the orthogonal projection matrix determination unit is used to determine the equivalent channel matrix of the user terminal according to the B_i obtained by the orthogonalization processing unit to orthographic projection matrix of the spanned subspace The simplified SVD algorithm determination unit is used to determine the value determined by the orthogonal projection matrix determination unit according to the number of receiving antennas of the user terminal As the matrix to be subjected to singular value decomposition SVD, determine the simplified SVD algorithm corresponding to the number of receiving antennas of the user terminal from the corresponding relationship between the preset number of receiving antennas and the simplified SVD algorithm for the matrix to be performed on the SVD; precoding Vector determining unit, for determining according to the simplified SVD algorithm determined based on the simplified SVD algorithm determining unit The right singular vector corresponding to the singular value of is used to determine the precoding vector for the user terminal.

一种基站，应用于TDD系统的单用户终端传输方案中，包括：矩阵与数量确定单元，用于确定所述用户终端对应的下行信道矩阵H_i和所述用户终端接收天线的数量；简化SVD算法确定单元，用于以矩阵与数量确定单元确定的Hi作为待进行SVD的矩阵，从预先设置的接收天线数量与针对待进行SVD的矩阵执行简化SVD算法的对应关系中，确定矩阵与数量确定单元确定的该数量所对应的简化SVD算法；预编码向量确定单元，用于根据简化SVD算法确定单元确定的简化SVD算法确定H_i的奇异值对应的右奇异向量，确定针对所述用户终端的预编码向量。A base station, applied to a single user terminal transmission scheme of a TDD system, comprising: a matrix and quantity determination unit, configured to determine the downlink channel matrix H_i corresponding to the user terminal and the number of receiving antennas of the user terminal; simplify SVD The algorithm determination unit is used to determine the matrix and the quantity from the corresponding relationship between the preset number of receiving antennas and the simplified SVD algorithm for the matrix to be performed SVD. The simplified SVD algorithm corresponding to the quantity determined by the unit; the precoding vector determination unit is used to determine the right singular vector corresponding to the singular value of H_i according to the simplified SVD algorithm determined by the simplified SVD algorithm determination unit, and determine the user terminal precoded vector.

一种用户终端，应用于FDD系统的单用户终端传输方案中，其特征在于，包括：矩阵与数量确定单元，用于确定所述用户终端的下行信道矩阵H_i，并确定所述用户终端接收天线的数量；简化SVD算法确定单元，用于以矩阵与数量确定单元确定的H_i作为待进行SVD的矩阵，从预先设置的接收天线数量与针对待进行SVD的矩阵执行简化SVD算法的对应关系中，确定与矩阵与数量确定单元确定的该数量对应的简化SVD算法；发送单元，用于根据简化SVD算法确定单元确定的简化SVD算法所确定的Hi的奇异值对应的右奇异向量，确定并向基站发送相应的PMI，指示基站根据PMI确定相应的预编码向量。A user terminal, which is applied to a single-user terminal transmission scheme of an FDD system, is characterized in that it includes: a matrix and quantity determination unit, configured to determine the downlink channel matrix H_i of the user terminal, and determine that the user terminal receives The number of antennas; the simplified SVD algorithm determination unit is used to determine the H_i determined by the matrix and the number determination unit as the matrix to be performed SVD, from the preset receiving antenna quantity and the corresponding relationship between the simplified SVD algorithm for the matrix to be performed SVD Among them, determine the simplified SVD algorithm corresponding to the quantity determined by the matrix and quantity determination unit; the sending unit is used to determine the right singular vector corresponding to the singular value of Hi determined by the simplified SVD algorithm determined by the simplified SVD algorithm determination unit, and determine and Send the corresponding PMI to the base station, and instruct the base station to determine the corresponding precoding vector according to the PMI.

本发明实施例的有益效果如下：The beneficial effects of the embodiments of the present invention are as follows:

通过本发明实施例提供的上述方法，在多用户终端传输方案中，无需再求取干扰信道矩阵零空间并将用户信道矩阵投影到该零空间，而是直接通过将用户信道矩阵正交投影到干扰信道矩阵的子空间，从而大大节约了运算资源。此外，简化SVD算法在本方案中的应用，也避免了利用现有技术提供的SVD算法需要执行多次迭代的缺陷，实现了对运算资源的节约。Through the above method provided by the embodiment of the present invention, in the multi-user terminal transmission scheme, it is no longer necessary to obtain the null space of the interference channel matrix and project the user channel matrix to the null space, but to directly project the user channel matrix onto the Interfering with the subspace of the channel matrix greatly saves computing resources. In addition, the application of the simplified SVD algorithm in this solution also avoids the defect that the SVD algorithm provided by the prior art needs to perform multiple iterations, and realizes the saving of computing resources.

附图说明Description of drawings

图1为本发明实施例提供的应用于多用户终端传输方案中的预编码向量的确定方法的具体流程示意图；FIG. 1 is a schematic flowchart of a method for determining a precoding vector applied to a multi-user terminal transmission scheme provided by an embodiment of the present invention;

图2为本发明实施例提供的一种应用于TDD系统的单用户终端传输方案中的预编码向量的确定方法的具体流程示意图；FIG. 2 is a schematic flowchart of a method for determining a precoding vector applied to a single-user terminal transmission scheme of a TDD system provided by an embodiment of the present invention;

图3为本发明实施例提供的一种应用于FDD系统的单用户终端传输方案中的预编码向量的确定方法的具体流程示意图；FIG. 3 is a schematic flowchart of a method for determining a precoding vector applied to a single-user terminal transmission scheme of an FDD system provided by an embodiment of the present invention;

图4为本发明实施例提供的应用于多用户终端传输方案中的预编码向量的确定装置的具体结构示意图；FIG. 4 is a schematic structural diagram of a device for determining a precoding vector applied to a multi-user terminal transmission scheme provided by an embodiment of the present invention;

图5为本发明实施例提供的一种应用于TDD系统的单用户终端传输方案中的基站的具体结构示意图；FIG. 5 is a schematic structural diagram of a base station in a single-user terminal transmission scheme applied to a TDD system provided by an embodiment of the present invention;

图6为本发明实施例提供的一种应用于FDD系统的单用户终端传输方案中的用户终端的具体结构示意图。FIG. 6 is a schematic structural diagram of a user terminal in a single-user terminal transmission solution applied to an FDD system provided by an embodiment of the present invention.

具体实施方式Detailed ways

为了解决现有技术提供的确定预编码向量的方式会消耗较多运算资源的问题，发明人对现有技术提供的预编码向量的确定方式进行了如下分析：In order to solve the problem that the method of determining the precoding vector provided by the prior art consumes more computing resources, the inventors analyzed the method of determining the precoding vector provided by the prior art as follows:

基于SVD分解的预编码是最优的，但典型的SVD求解算法(通常需要迭代)的运算量随着矩阵维数的增加而增加。在发送天线数较大但传输流数较少的情况下(在LTE Rel.8/9/10规范中，特别定义了单流和双流传输模式，而且在多用户终端传输模式下，每个UE传输流数不超过2)，完全SVD分解是没有必要的，因为此时只需要最大的n个奇异值对应的右奇异向量即可。在增强型多天线系统中，如果用户终端接收天线数配置为1或2，则不论发送端天线数如何，SVD都会存在闭式表达式，即用户MIMO信道矩阵H的右奇异向量可以由公式直接给出，而不需要迭代计算。同样，在传输流数较少的情况下(例如单流或双流)，SVD分解同样存在精确的近似解。Precoding based on SVD decomposition is optimal, but the computational complexity of typical SVD solving algorithms (usually requiring iterations) increases as the matrix dimension increases. In the case of a large number of transmit antennas but a small number of transmission streams (in the LTE Rel. The number of transmission streams does not exceed 2), and complete SVD decomposition is not necessary, because at this time only the right singular vector corresponding to the largest n singular values is required. In the enhanced multi-antenna system, if the number of receiving antennas of the user terminal is configured as 1 or 2, then regardless of the number of antennas at the transmitting end, SVD will have a closed-form expression, that is, the right singular vector of the user MIMO channel matrix H can be directly expressed by the formula given without iterative computation. Similarly, in the case of a small number of transmission streams (such as single stream or dual stream), the SVD decomposition also has an accurate approximate solution.

同样，干扰信道的零空间的求解运算量也随着干扰信道维数的增加而增加，这在多达24根发送天线的CoMP系统中，是很难实现的。事实上，多用户预编码的迫零处理，其本质是获得用户信道矩阵到干扰信道矩阵的正交投影。正交投影既可以通过求取干扰信道矩阵零空间，然后将用户信道矩阵投影到该零空间实现，也可以直接通过将用户信道矩阵正交投影到干扰信道矩阵的子空间实现，显然，后者在下行传输流数较少(相对发送天线数)的情况下，运算量较前者低得多。Similarly, the amount of computation for solving the null space of the interference channel also increases with the increase of the dimension of the interference channel, which is difficult to achieve in a CoMP system with up to 24 transmit antennas. In fact, the essence of zero-forcing processing in multi-user precoding is to obtain the orthogonal projection from the user channel matrix to the interference channel matrix. Orthogonal projection can be achieved by calculating the null space of the interference channel matrix and then projecting the user channel matrix to the null space, or directly by orthogonally projecting the user channel matrix to the subspace of the interference channel matrix. Obviously, the latter When the number of downlink transmission streams is small (compared to the number of transmitting antennas), the calculation load is much lower than the former.

通过上述分析，发明人认为，事实上，在某些典型场景下，例如用户终端仅配置1或2根接收天线的场景下，SVD分解是可以简化的，即存在闭式解，在单流或双流传输以及用户终端配置4根接收天线情况下，SVD分解同样存在近似解。而且，不管采用何种SVD分解方法，多用户预编码中为消除干扰而进行的零空间求解都是可以避免的。Through the above analysis, the inventor believes that, in fact, in some typical scenarios, such as the scenario where the user terminal is only equipped with 1 or 2 receiving antennas, the SVD decomposition can be simplified, that is, there is a closed-form solution. In the case of dual-stream transmission and user terminal configuration with 4 receiving antennas, the SVD decomposition also has an approximate solution. Moreover, no matter what kind of SVD decomposition method is adopted, the null space solution for interference elimination in multi-user precoding can be avoided.

鉴于此，针对用户终端接收不超过2流的传输情况，本发明提供了一种新的下行多天线系统的预编码方法，在该方法中，SVD分解直接由公式给出，多用户预编码的迫零处理通过正交投影获得，无需求解干扰信道的零空间。In view of this, the present invention provides a new precoding method for a downlink multi-antenna system for a user terminal receiving no more than 2 streams. In this method, the SVD decomposition is directly given by the formula, and the multi-user precoding The zero-forcing process is obtained by orthogonal projection, and there is no need to solve the null space of the interfering channel.

以下结合附图，对本发明实施例提供的方案进行详细说明。The solutions provided by the embodiments of the present invention will be described in detail below in conjunction with the accompanying drawings.

本发明实施例首先提供一种应用于多用户终端传输方案中的预编码向量的确定方法，该方法首先分别确定各个用户终端对应的等效信道矩阵T_j，然后，针对每一用户终端，通过执行如图1所示的以下步骤，实现确定针对该用户终端的预编码向量：Embodiments of the present invention firstly provide a method for determining a precoding vector applied to a multi-user terminal transmission scheme. The method first determines the equivalent channel matrix T_j corresponding to each user terminal, and then, for each user terminal, through Execute the following steps as shown in Figure 1 to realize determining the precoding vector for the user terminal:

步骤11，根据除该用户终端外的其他用户终端所对应的T_j，确定该用户终端的干扰信道矩阵Step 11, according to T_j corresponding to other user terminals except the user terminal, determine the interference channel matrix of the user terminal

步骤12，对中的所有行向量执行正交化处理，获得到所张成的子空间的一组标准正交基Bi，其中，这里正交化处理可以但不限于采用斯密特算法(包括改进的斯密特算法)、行列初等变换算法、Householder变换算法以及Givens旋转等算法来实现；Step 12, yes Orthogonalization is performed on all the row vectors in to get A group of orthonormal bases Bi of the subspace that is stretched, wherein, the orthogonalization process here can be but not limited to adopt Schmidt algorithm (including improved Schmidt algorithm), rank and column elementary transformation algorithm, Householder transformation algorithm and Givens rotation and other algorithms to achieve;

步骤13，根据B_i确定该用户终端的等效信道矩阵到所张成的子空间的正交投影矩阵Step 13, determine the equivalent channel matrix of the user terminal according to_Bi to orthographic projection matrix of the spanned subspace

步骤14，确定该用户终端接收天线的数量，并以作为待进行奇异值分解SVD的矩阵，从预先设置的接收天线数量与针对待进行SVD的矩阵执行简化SVD算法的对应关系中，确定与确定的该数量对应的简化SVD算法；Step 14, determine the number of receiving antennas of the user terminal, and use As the matrix to be subjected to singular value decomposition SVD, determine the simplified SVD algorithm corresponding to the determined quantity from the corresponding relationship between the preset number of receiving antennas and the simplified SVD algorithm for the matrix to be performed on SVD;

步骤15，根据确定的简化SVD算法确定的的奇异值对应的右奇异向量，确定针对该用户终端的预编码向量。在本步骤15中，根据的奇异值对应的右奇异向量确定针对该用户终端的预编码向量时，可以先对该用户终端的传输方式为进行判断，当判断出该传输方式为单流传输时，将的最大奇异值对应的右奇异向量确定为该用户终端对应的预编码向量，而当判断出该传输方式为双流传输时，根据的最大和第二大奇异值分别对应的右奇异向量确定该预编码向量。Step 15, determined according to the determined simplified SVD algorithm The right singular vector corresponding to the singular value of is used to determine the precoding vector for the user terminal. In this step 15, according to When the right singular vector corresponding to the singular value of is determined as the precoding vector for the user terminal, the transmission mode of the user terminal can be judged first, and when it is judged that the transmission mode is single-stream transmission, the The right singular vector corresponding to the largest singular value of is determined as the precoding vector corresponding to the user terminal, and when it is judged that the transmission mode is dual-stream transmission, according to The right singular vectors corresponding to the largest and second largest singular values of , determine the precoding vector.

通过本发明实施例提供的上述方法，在多用户终端传输方案中，无需再求取干扰信道矩阵零空间并将用户信道矩阵投影到该零空间，而是直接通过将用户信道矩阵正交投影到干扰信道矩阵的子空间，从而大大节约了运算资源。Through the above method provided by the embodiment of the present invention, in the multi-user terminal transmission scheme, it is no longer necessary to obtain the null space of the interference channel matrix and project the user channel matrix to the null space, but directly by orthogonally projecting the user channel matrix to Interfering with the subspace of the channel matrix greatly saves computing resources.

此外，上述简化SVD算法的应用，也避免了利用现有技术提供的SVD算法需要执行多次迭代的缺陷，从而也实现了运算资源的节约。具体地，上述预先设置的接收天线数量与简化SVD算法的对应关系具体为：In addition, the application of the above-mentioned simplified SVD algorithm also avoids the defect that the SVD algorithm provided by the prior art needs to perform multiple iterations, thereby realizing the saving of computing resources. Specifically, the corresponding relationship between the preset number of receiving antennas and the simplified SVD algorithm is specifically:

1、当接收天线的数量为1时，相应的简化SVD算法为：以|H|作为对待进行SVD的矩阵H进行SVD后得到的最大奇异值，并以H|H|^-1作为对应于|H|的右奇异值向量，其中，|H|表示计算H的2-范数；1. When the number of receiving antennas is 1, the corresponding simplified SVD algorithm is: take |H^| The right singular value vector of H|, where |H| means calculating the 2-norm of H;

2、当接收天线的数量为2时，相应的简化SVD算法为：以λ₁²和分别作为对H进行SVD后得到的最大奇异值和第二大奇异值，并以V₁和V₂分别作为对应于λ₁²和的右奇异值向量，其中，按照下述公式[1]计算λ₁²：2. When the number of receiving antennas is 2, the corresponding simplified SVD algorithm is: with λ₁² and Respectively as the largest singular value and the second largest singular value obtained after SVD on H, and take V₁ and V₂ as corresponding to λ₁² and The right singular value vector of , where λ₁² is calculated according to the following formula [1]:

${λ λ}_{11}^{22} = = \frac{{| | {h h}_{11} | |}^{22} {((β β + + \sqrt{11 - - {β β}^{22}} | | η η | |))}^{22} + + {| | {h h}_{22} | |}^{22} {((β β | | η η | | + + \sqrt{11 - - {β β}^{22}}))}^{22}}{11 + + 22 β β \sqrt{11 - - {β β}^{22}} | | η η | |} - - - - - - [[11]]$

按照下述公式[2]计算Calculate according to the following formula [2]

${λ λ}_{22}^{22} = = {| | {h h}_{11} | |}^{22} + + {| | {h h}_{22} | |}^{22} - - {λ λ}_{11}^{22} - - - - - - [[22]]$

按照下述公式计算V₁：Calculate V₁ according to the following formula:

${V V}_{11} = = \frac{β β {h h}_{11} {| | {h h}_{11} | |}^{- - 11} + + \sqrt{11 - - {β β}^{22}} η η {| | η η | |}^{- - 11} {h h}_{22} {| | {h h}_{22} | |}^{- - 11}}{\sqrt{11 + + 22 β β \sqrt{11 - - {β β}^{22}} | | η η | |}} - - - - - - [[33]]$

按照下述公式计算V₂：Calculate V₂ according to the following formula:

${V V}_{22} = = \frac{{h h}_{11} - - {h h}_{11} {V V}_{11}^{H h} {V V}_{11}}{| | {h h}_{11} - - {h h}_{11} {V V}_{11}^{H h} {V V}_{11} | |} - - - - - - [[44]]$

其中，h₁为H的第一个行向量，h₂为H的第二个行向量，上述公式中的η和β分别按照下述公式[5]、[6]计算：Among them, h₁ is the first row vector of H, h₂ is the second row vector of H, η and β in the above formula are calculated according to the following formulas [5] and [6] respectively:

$η η = = {h h}_{11} {h h}_{22}^{H h} {| | {h h}_{11} | |}^{- - 11} {| | {h h}_{22} | |}^{- - 11} - - - - - - [[55]]$

$β β = = \sqrt{\frac{11}{22} \frac{22 {| | {h h}_{11} | |}^{22} (({| | {h h}_{11} | |}^{22} + + {| | {h h}_{22} | |}^{22})) {| | η η | |}^{22} + + {(({| | {h h}_{11} | |}^{22} - - {| | {h h}_{22} | |}^{22}))}^{22} + + (({| | {h h}_{11} | |}^{22} - - {| | {h h}_{22} | |}^{22})) \sqrt{{(({| | {h h}_{11} | |}^{22} - - {| | {h h}_{22} | |}^{22}))}^{22} + + 44 {| | {h h}_{11} | |}^{22} {| | {h h}_{22} | |}^{22} {| | η η | |}^{22}}}{{(({| | {h h}_{11} | |}^{22} - - {| | {h h}_{22} | |}^{22}))}^{22} + + {(({| | {h h}_{11} | |}^{22} + + {| | {h h}_{22} | |}^{22}))}^{22} {| | η η | |}^{22}}} - - - - - - [[66]]$

3、当接收天线的数量为4时，相应的简化SVD算法为：以矩阵作为待进行SVD的矩阵，按照为2的接收天线数量对应的简化SVD算法，分别确定的最大奇异值、第二大奇异值、最大奇异值对应的右奇异向量、第二大奇异值对应的右奇异向量，作为H相应的最大奇异值、第二大奇异值、最大奇异值对应的右奇异向量、第二大奇异值对应的右奇异向量；其中，λ_1，1、V_1，1、λ_2，1、V_2，1按照下述方式确定：3. When the number of receiving antennas is 4, the corresponding simplified SVD algorithm is: the matrix As the matrix to be SVD, according to the simplified SVD algorithm corresponding to the number of receiving antennas of 2, respectively determine The largest singular value, the second largest singular value, the right singular vector corresponding to the largest singular value, and the right singular vector corresponding to the second largest singular value, as the corresponding largest singular value, second largest singular value, and largest singular value of H The right singular vector and the right singular vector corresponding to the second largest singular value; among them, λ_1,1 , V_1,1 , λ_2,1 , V_2,1 are determined as follows:

对H进行划分得到H₁和H₂，其中H₁由H的两个行向量组成，H₂由H的另外两个行向量组成；以H₁和H₂作为待进行SVD的矩阵，按照的接收天线数量为2时所对应的简化SVD算法，分别确定H₁的最大奇异值λ_1，1与相应的右奇异向量V_1，1，以及H₂的最大奇异值λ_2，1与相应的右奇异向量V_2，1。Divide H to obtain H₁ and H₂ , where H₁ is composed of two row vectors of H, and H₂ is composed of the other two row vectors of H; H₁ and H₂ are used as the matrix to be SVD, according to The corresponding simplified SVD algorithm when the number of receiving antennas is 2, respectively determine the maximum singular value λ_1,1 of H₁ and the corresponding right singular vector V_1,1 , and the maximum singular value λ_2,1 of H₂ and the corresponding Right singular vector V_2,1 .

需要说明的是，本发明实施例提供的该方法分别应用于TDD系统和FDD时，各个用户终端对应的等效信道矩阵T_j的确定方式上存在一些差异。具体差异如下：It should be noted that when the method provided by the embodiment of the present invention is applied to the TDD system and the FDD system respectively, there are some differences in the way of determining the equivalent channel matrix T_j corresponding to each user terminal. The specific differences are as follows:

当该方法应用于TDD系统时，确定各个用户终端对应的等效信道矩阵T_j的方式为：首先，根据各个用户终端发送的导频信号SRS确定各个用户终端对应的下行信道矩阵H_j；然后，分别判断针对各个用户终端的传输方式为单流传输或双流传输；并针对判断出传输方式为单流传输的每一单流传输用户终端执行：确定该单流传输用户终端接收天线的数量，并以H_j作为待进行SVD的矩阵，从上文中所述的对应关系中，确定与该单流传输用户终端接收天线的数量对应的简化SVD算法；根据确定的简化SVD算法确定H_j的最大奇异值对应的右奇异向量，并将该最大奇异值对应的右奇异向量确定为该单流传输用户终端的T_j；而针对判断出传输方式为双流传输的每一双流传输用户终端，直接将该双流传输用户终端的H_j确定为该双流传输用户终端的T_j。When this method is applied to a TDD system, the way to determine the equivalent channel matrix T_j corresponding to each user terminal is: first, determine the downlink channel matrix H_j corresponding to each user terminal according to the pilot signal SRS sent by each user terminal; then , respectively determine whether the transmission mode for each user terminal is single-stream transmission or dual-stream transmission; and for each single-stream transmission user terminal determined to be single-stream transmission, perform: determine the number of receiving antennas of the single-stream transmission user terminal, And use H_j as the matrix to be performed SVD, from the above-mentioned corresponding relationship, determine the simplified SVD algorithm corresponding to the number of receiving antennas of the user terminal for this single-stream transmission; determine the maximum value of H_j according to the determined simplified SVD algorithm The right singular vector corresponding to the singular value, and determine the right singular vector corresponding to the largest singular value as the T_j of the single-stream transmission user terminal; and for each dual-stream transmission user terminal whose transmission mode is determined to be dual-stream transmission, directly H_j of the dual-stream transmission user terminal is determined as T_j of the dual-stream transmission user terminal.

当该方法应用于FDD系统时，确定各个用户终端对应的等效信道矩阵T_j的方式为：首先，获得各个用户终端根据基站发送的下行公共导频信号CRS反馈的码字索引PMI；然后，分别判断针对各个用户终端的传输方式为单流传输或双流传输；并针对判断出传输方式为单流传输的每一单流传输用户终端，将该单流传输用户终端反馈的PMI对应的预编码码字的第一列向量确定为该单流传输用户终端的T_j；而针对判断出所述传输方式为双流传输的每一双流传输用户终端，根据该双流传输用户终端反馈的PMI对应的预编码码字的前两列向量确定该双流传输用户终端的T_j。When this method is applied to the FDD system, the way to determine the equivalent channel matrix T_j corresponding to each user terminal is: first, obtain the codeword index PMI fed back by each user terminal according to the downlink common pilot signal CRS sent by the base station; then, Respectively determine whether the transmission mode for each user terminal is single-stream transmission or dual-stream transmission; and for each single-stream transmission user terminal for which the transmission mode is determined to be single-stream transmission, precoding corresponding to the PMI fed back by the single-stream transmission user terminal The first column vector of the codeword is determined to be the T_j of the single-stream transmission user terminal; and for each dual-stream transmission user terminal whose transmission mode is determined to be dual-stream transmission, according to the prediction corresponding to the PMI fed back by the dual-stream transmission user terminal The first two column vectors of the encoded codeword determine the T_j of the dual-stream transmission user terminal.

此外，本发明实施例还提供一种应用于TDD系统的单用户终端传输方案中的预编码向量的确定方法，该方法包括如图2所示的下述步骤：In addition, an embodiment of the present invention also provides a method for determining a precoding vector applied to a single-user terminal transmission scheme of a TDD system, the method including the following steps as shown in FIG. 2:

步骤21，基站确定用户终端对应的下行信道矩阵H_i，并确定该用户终端接收天线的数量；Step 21, the base station determines the downlink channel matrix H_i corresponding to the user terminal, and determines the number of receiving antennas of the user terminal;

步骤22，以H_i作为待进行SVD的矩阵，从预先设置的接收天线数量与针对待进行SVD的矩阵执行简化SVD算法的上述对应关系中，确定与确定的该数量对应的简化SVD算法；Step 22, using H_i as the matrix to be SVD, from the above-mentioned corresponding relationship between the preset number of receiving antennas and the simplified SVD algorithm for the matrix to be SVD, determine the simplified SVD algorithm corresponding to the determined number;

步骤23，根据确定的简化SVD算法确定H_i的奇异值对应的右奇异向量，确定针对该用户终端的预编码向量。具体地，可以先对该用户终端的传输方式进行判断，当判断出该传输方式为单流传输时，将H_i的最大奇异值对应的右奇异向量确定为该用户终端对应的预编码向量，而当判断出该传输方式为双流传输时，根据H_i的最大和第二大奇异值分别对应的右奇异向量确定该用户终端对应的预编码向量。Step 23: Determine the right singular vector corresponding to the singular value of H_i according to the determined simplified SVD algorithm, and determine the precoding vector for the user terminal. Specifically, the transmission mode of the user terminal may be judged first, and when it is judged that the transmission mode is single-stream transmission, the right singular vector corresponding to the largest singular value of H_i is determined as the precoding vector corresponding to the user terminal, When it is determined that the transmission mode is dual-stream transmission, the precoding vector corresponding to the user terminal is determined according to the right singular vectors corresponding to the largest and second largest singular values of H_i respectively.

由上述应用在TDD系统单用户终端传输方案中的预编码向量的确定方法可知，由于无需通过迭代的方式对H_i进行SVD，而是只需利用预设的简化SVD算法来方便地对H_i进行SVD，从而大大节省了运算资源。From the above method of determining the precoding vector applied in the single-user terminal transmission scheme of the TDD system, it can be seen that since it is not necessary to perform SVD on H_i in an iterative manner, it is only necessary to use the preset simplified SVD algorithm to conveniently perform SVD on H_i SVD is performed, thereby greatly saving computing resources.

此外，本发明实施例还提供一种预编码向量的确定方法，应用于FDD系统的单用户终端传输方案中，该方法包括如图3所示的下述步骤：In addition, an embodiment of the present invention also provides a method for determining a precoding vector, which is applied to a single-user terminal transmission scheme of an FDD system, and the method includes the following steps as shown in FIG. 3 :

步骤31，用户终端确定自身的下行信道矩阵H_i，并确定自身接收天线的数量；Step 31, the user terminal determines its own downlink channel matrix H_i , and determines the number of its own receiving antennas;

步骤32，以H_i作为待进行SVD的矩阵，从预先设置的接收天线数量与针对待进行SVD的矩阵执行简化SVD算法的上述对应关系中，确定与确定的该数量对应的简化SVD算法；Step 32, using H_i as the matrix to be SVD, from the above-mentioned corresponding relationship between the preset number of receiving antennas and the simplified SVD algorithm for the matrix to be SVD, determine the simplified SVD algorithm corresponding to the determined number;

步骤33，根据确定的简化SVD算法确定H_i的奇异值对应的右奇异向量，并根据右奇异向量确定并向基站发送相应的PMI，指示基站根据PMI确定相应的预编码向量。类似地，用户终端也可以在判断出所用的传输方式为单流传输时，从预编码码本中，选取相应于H_i的最大奇异值所对应的右奇异向量的码字，并确定与发送选取的该码字对应的PMI；而在判断出传输方式为双流传输时，从预编码码本中，选取相应于H_i的最大和第二大奇异值分所别对应的右奇异向量构成的右奇异矩阵的码字，并确定与发送选取的该码字对应的PMI。Step 33: Determine the right singular vector corresponding to the singular value of_Hi according to the determined simplified SVD algorithm, determine and send the corresponding PMI to the base station according to the right singular value, and instruct the base station to determine the corresponding precoding vector according to the PMI. Similarly, when judging that the transmission mode used is single-stream transmission, the user terminal may select the codeword corresponding to the right singular vector corresponding to the largest singular value of H_i from the precoding codebook, and determine and send The PMI corresponding to the selected codeword; and when it is judged that the transmission mode is dual-stream transmission, from the precoding codebook, select the corresponding right singular vector corresponding to the largest and second largest singular value of H_i respectively. The codeword of the right singular matrix, and determine the PMI corresponding to the codeword selected by the transmission.

基于上述方法，对本发明实施例提供的方案在实际中的应用的详细介绍如下：Based on the above method, the detailed introduction to the practical application of the solution provided by the embodiment of the present invention is as follows:

步骤一：估计用户终端下行信道矩阵H并执行SVD。Step 1: Estimate the user terminal downlink channel matrix H and perform SVD.

在TDD系统中，由基站端估计H并对H执行SVD。H的估计与SVD方式如下：In a TDD system, H is estimated by the base station and SVD is performed on H. The estimation of H and the SVD method are as follows:

1、用户端发送探测导频信号SRS；1. The user end sends a sounding pilot signal SRS;

2、基站端利用用户端发送的SRS和信道互易性估计H；2. The base station uses the SRS sent by the user terminal and the channel reciprocity to estimate H;

3、基站端对H执行SVD。3. The base station performs SVD on H.

在FDD系统中，由用户端估计H并对H执行SVD。H的估计与SVD方式如下：In FDD systems, H is estimated by the UE and SVD is performed on H. The estimation of H and the SVD method are as follows:

1、基站端发送下行公共导频信号CRS；1. The base station sends the downlink common pilot signal CRS;

2、用户端利用基站端发送的CRS估计H；2. The UE uses the CRS sent by the base station to estimate H;

3、用户端对H执行SVD得到H的右奇异向量，并根据右奇异向量从预编码码本中选择一个码字，并反馈该码字的索引PMI到基站端。3. The UE performs SVD on H to obtain the right singular vector of H, selects a codeword from the precoding codebook according to the right singular vector, and feeds back the index PMI of the codeword to the base station.

在本发明实施例中，记需要执行SVD的矩阵为H，一种简化的SVD分解方法如下：In the embodiment of the present invention, remember that the matrix that needs to perform SVD is H, and a simplified SVD decomposition method is as follows:

1)如果H为1×N维矩阵，则其最大奇异值为λ＝|H|，对应的奇异值向量为V＝H|H|^-1，其中|H|表示计算H的2-范数，N为基站端发送天线数，取值为大于等于2且不大于24的偶数；1) If H is a 1×N-dimensional matrix, its maximum singular value is λ=|H|, and the corresponding singular value vector is V=H|H|^-1 , where |H| represents the calculation of the 2-norm of H , N is the number of transmitting antennas at the base station, and the value is an even number greater than or equal to 2 and not greater than 24;

2)如果H为2×N维矩阵，记第1个行向量为h₁，第2个行向量为h₂，则H的最大奇异值为对应的奇异值向量为其中以及第二大特征值为对应的奇异值向量为2) If H is a 2×N-dimensional matrix, record the first row vector as h₁ and the second row vector as h₂ , then the maximum singular value of H is The corresponding singular value vector is in as well as The second largest eigenvalue is The corresponding singular value vector is

3)如果H为4×N维矩阵，执行如下步骤，得到H的SVD分解近似值：3) If H is a 4×N-dimensional matrix, perform the following steps to obtain the approximate value of the SVD decomposition of H:

首先，将H分为两个2×N维矩阵H₁和H₂，其中H₁由H的两个行向量组成，H₂由H的另外两个行向量组成；First, divide H into two 2×N-dimensional matrices H₁ and H₂ , where H₁ consists of two row vectors of H, and H₂ consists of the other two row vectors of H;

然后，根据上述2)的方式，对H₁和H₂执行SVD，得到H₁的最大奇异值λ_1，1以及对应的右奇异向量V_1，1，得到H₂的最大奇异值λ_2，1以及对应的右奇异向量V_2，1；Then, according to the method of 2) above, perform SVD on H₁ and H₂ to obtain the maximum singular value λ_1,1 of H₁ and the corresponding right singular vector V_1,1 , and obtain the maximum singular value λ 2 of H₂_{, 1} and the corresponding right singular vector V_2,1 ;

最后，根据步骤上述2)的方式，对矩阵执行SVD，得到最大奇异值λ₁以及对应的右奇异向量V₁，以及次最大奇异值λ₂以及对应的右奇异向量V₂，λ₁和λ₂以及对应的V₁和V₂即是H的SVD近似解。Finally, according to the method of step 2) above, the matrix Execute SVD to get the largest singular value λ₁ and the corresponding right singular vector V₁ , and the second largest singular value λ₂ and the corresponding right singular vector V₂ , λ₁ and λ₂ and the corresponding V₁ and V₂ are H The SVD approximate solution of .

步骤二：计算用户终端i的预编码向量。Step 2: Calculate the precoding vector of user terminal i.

在多用户终端传输方案中，基站计算用户终端i的预编码向量过程如下：In the multi-user terminal transmission scheme, the base station calculates the precoding vector of user terminal i as follows:

首先，记用户终端i的干扰信道矩阵为其中M为同传用户终端数，T_j为用户终端j(j＝1，2，...，M)的等效信道矩阵；First, record the interference channel matrix of user terminal i as Wherein M is the number of simultaneous interpretation user terminals, and T_j is the equivalent channel matrix of user terminal j (j=1,2,...,M);

在TDD系统中，若用户终端j为单流传输方式，则T_j＝V_1，j，这里V_1，j为用户终端MIMO信道矩阵H_j的最大奇异值对应的右奇异向量，根据上述2)的方式对H_j执行简化SVD获得，若用户终端j是双流传输方式，则T_j＝H_j。In the TDD system, if the user terminal j is in single-stream transmission mode, then T_j =V_1,j , where V_1,j is the right singular vector corresponding to the largest singular value of the user terminal MIMO channel matrix H_j , according to the above 2 ) to obtain H_j through simplified SVD, if user terminal j is in dual-stream transmission mode, then T_j =H_j .

在FDD系统中，若用户j终端为单流传输方式，T_j为用户终端j反馈的PMI对应的预编码码字的第一列向量，若用户终端j为双流传输方式，则T_j为用户终端j反馈的PMI对应的预编码码字的前两列向量。In the FDD system, if the user terminal j is in the single-stream transmission mode, T_j is the first column vector of the precoding codeword corresponding to the PMI fed back by the user terminal j, and if the user terminal j is in the dual-stream transmission mode, then T_j is the user The first two column vectors of the precoding codeword corresponding to the PMI fed back by terminal j.

然后，对中所有行向量执行正交化方法，得到所张成的子空间的一组标准正交基其中L为矩阵的行向量个数；then, yes Performing the orthogonalization method on all row vectors in , we get A set of orthonormal basis for the spanned subspace where L is the matrix The number of row vectors;

接着，计算用户终端i等效信道矩阵T_i到所张成的子空间的正交投影即Next, calculate the equivalent channel matrix T_i of user terminal i to Orthographic projection of the spanned subspace Right now

最后，根据确定用户终端i的预编码向量。Finally, according to Determine the precoding vector for user terminal i.

若用户终端i采用的是单流传输方式，则用户终端i的预编码向量为其中V_1，i为最大奇异值对应的右奇异向量，该右奇异向量通过采用上述2)的方式对执行简化SVD获得；若户终端i采用的是双流传输方式，则用户终端i的预编码向量为其中V_1，i和V_2，i分别为最大和第二大奇异值对应的右奇异向量，该右奇异向量通过采用上述2)的方式对执行简化SVD获得。If user terminal i adopts single-stream transmission mode, the precoding vector of user terminal i is where_V1,i is The right singular vector corresponding to the largest singular value, the right singular vector is corrected by adopting the above 2) method Perform simplified SVD to obtain; if user terminal i adopts dual-stream transmission mode, then the precoding vector of user terminal i is where V_1,i and V_2,i are respectively The right singular vectors corresponding to the largest and second largest singular values, the right singular vectors are paired by using the above 2) method Perform simplified SVD to obtain.

在本发明实施例中，针对单用户终端传输的方案，基站计算用户终端i的预编码向量过程如下：In the embodiment of the present invention, for the single-user terminal transmission scheme, the base station calculates the precoding vector of user terminal i as follows:

在TDD系统中，若是用户终端i采用的单流传输方式，则用户终端i的预编码向量为其中V_1，i为用户i的信道矩阵H_i最大奇异值对应的右奇异向量，该右奇异向量通过采用上述2)的方式对H_i执行简化SVD获得；若是双流传输方式，则用户终端i的预编码向量为其中V_1，i和V_2，i分别为用户终端i的信道矩阵H_i最大和第二大奇异值对应的右奇异向量，该右奇异向量通过采用上述2)的方式对H_i执行简化SVD获得。In a TDD system, if user terminal i adopts a single-stream transmission mode, then the precoding vector of user terminal i is Where V_{1, i} is the right singular vector corresponding to the largest singular value of user i's channel matrix H_i , the right singular vector is obtained by performing simplified SVD on H_i using the method of 2) above; if it is a dual-stream transmission mode, then user terminal i The precoding vector of is where V_{1, i} and V_{2, i} are the right singular vectors corresponding to the largest and second largest singular values of the channel matrix H_i of user terminal i respectively, and the right singular vector performs simplified SVD on H_i by adopting the method of 2) above get.

在FDD系统中，若是用户终端i采用的单流传输方式，则用户终端i通过采用上述2)中的方式对H_i执行简化SVD获得最大奇异值对应的右奇异向量V_1，i，并根据V_1，i从预编码码本中选择一个码字，并反馈该码字的索引PMI到基站，则用户终端i的预编码向量为用户i反馈的PMI对应的预编码码字；若是双流传输，用户终端i通过上述方式2)对H_i执行简化SVD获得最大奇异值和第二大奇异值对应的右奇异矩阵根据从预编码码本中选择一个码字，并反馈该码字的索引PMI到基站，则用户终端i的预编码向量为用户终端i反馈的PMI对应的预编码码字。In the FDD system, if user terminal i adopts a single-stream transmission method, user terminal i performs simplified SVD on H_i by adopting the method in 2) above to obtain the right singular vector V_1,i corresponding to the largest singular value, and according to V_{1, i} selects a codeword from the precoding codebook, and feeds back the index PMI of the codeword to the base station, then the precoding vector of user terminal i is the precoding codeword corresponding to the PMI fed back by user i; if dual-stream transmission , user terminal i obtains the right singular matrix corresponding to the largest singular value and the second largest singular value by performing simplified SVD on H_i through the above method 2) according to Select a codeword from the precoding codebook, and feed back the index PMI of the codeword to the base station, then the precoding vector of user terminal i is the precoding codeword corresponding to the PMI fed back by user terminal i.

对应于本发明实施例提供的预编码向量的确定方法，本发明实施例还提供一种预编码向量的确定装置，该装置的具体结构示意图如图4所示，包括以下功能单元：Corresponding to the method for determining the precoding vector provided by the embodiment of the present invention, the embodiment of the present invention also provides a device for determining the precoding vector. The specific structural diagram of the device is shown in Figure 4, including the following functional units:

等效信道矩阵确定单元41，用于确定各个用户终端对应的等效信道矩阵T_j；An equivalent channel matrix determining unit 41, configured to determine an equivalent channel matrix T_j corresponding to each user terminal;

干扰信道矩阵确定单元42，用于针对每一用户终端，根据等效信道矩阵确定单元41确定的除该用户终端外的其他用户终端所对应的等效信道矩阵T_j，确定该用户终端的干扰信道矩阵The interference channel matrix determination unit 42 is configured to, for each user terminal, determine the interference of the user terminal according to the equivalent channel matrix T_j corresponding to other user terminals other than the user terminal determined by the equivalent channel matrix determination unit 41 channel matrix

正交化处理单元43，用于对干扰信道矩阵确定单元42确定的中的所有行向量执行正交化处理，得到所张成的子空间的一组标准正交基B_i；Orthogonalization processing unit 43, for determining the interference channel matrix determining unit 42 Orthogonalization is performed on all the row vectors in to get A set of orthonormal basis B_i of the spanned subspace;

正交投影矩阵确定单元44，用于根据正交化处理单元43得到的B_i，确定该用户终端的等效信道矩阵到所张成的子空间的正交投影矩阵Orthogonal projection matrix determination unit 44, configured to determine the equivalent channel matrix of the user terminal according to the B_i obtained by the orthogonalization processing unit 43 to orthographic projection matrix of the spanned subspace

简化SVD算法确定单元45，用于根据该用户终端接收天线的数量，以正交投影矩阵确定单元44确定的作为待进行奇异值分解SVD的矩阵，从预先设置的接收天线数量与针对待进行SVD的矩阵执行简化SVD算法的对应关系中，确定与该用户终端接收天线的数量对应的简化SVD算法；The simplified SVD algorithm determining unit 45 is used to determine the value determined by the orthogonal projection matrix determining unit 44 according to the number of receiving antennas of the user terminal. As a matrix to be subjected to singular value decomposition SVD, determine a simplified SVD algorithm corresponding to the number of receiving antennas of the user terminal from the corresponding relationship between the preset number of receiving antennas and the simplified SVD algorithm for the matrix to be performed on SVD;

预编码向量确定单元46，用于根据基于简化SVD算法确定单元45确定的简化SVD算法所确定的的奇异值对应的右奇异向量，确定针对该用户终端的预编码向量。A precoding vector determining unit 46, configured to determine the vector based on the simplified SVD algorithm determined by the simplified SVD algorithm determining unit 45 The right singular vector corresponding to the singular value of is used to determine the precoding vector for the user terminal.

当该装置应用于TDD系统时，上述等效信道矩阵确定单元41具体可以划分为以下功能模块：When the device is applied to a TDD system, the above-mentioned equivalent channel matrix determining unit 41 can be specifically divided into the following functional modules:

下行信道矩阵确定模块，用于根据各个用户终端发送的导频信号SRS，确定各个用户终端对应的下行信道矩阵H_j；第一判断模块，用于分别判断针对各个用户终端的传输方式为单流传输或双流传输；第一执行模块，用于针对第一判断模块判断出传输方式为单流传输的每一单流传输用户终端，执行：确定该单流传输用户终端接收天线的数量，并以下行信道矩阵确定模块确定的该单流传输用户终端的H_j作为待进行SVD的矩阵，从所述对应关系中，确定与该单流传输用户终端接收天线的数量对应的简化SVD算法；根据确定的简化SVD算法确定该单流传输用户终端的H_j的最大奇异值对应的右奇异向量，并将所述最大奇异值对应的右奇异向量确定为该单流传输用户终端的T_j；第二执行模块，用于针对第一判断模块判断出传输方式为双流传输的每一双流传输用户终端，执行：将该双流传输用户终端的H_j确定为该双流传输用户终端的T_j。The downlink channel matrix determination module is used to determine the downlink channel matrix H_j corresponding to each user terminal according to the pilot signal SRS sent by each user terminal; the first judgment module is used to judge that the transmission mode for each user terminal is single stream transmission or dual-stream transmission; the first execution module is configured to execute, for each single-stream transmission user terminal determined by the first judging module that the transmission mode is single-stream transmission: to determine the number of receiving antennas of the single-stream transmission user terminal, and perform the following The_Hj of the single-stream transmission user terminal determined by the row channel matrix determination module is used as the matrix to be performed SVD, and from the corresponding relationship, determine the simplified SVD algorithm corresponding to the number of receiving antennas of the single-stream transmission user terminal; according to the determination The simplified SVD algorithm of the single-stream transmission user terminal determines the right singular vector corresponding to the maximum singular value of H_j , and determines the right singular vector corresponding to the maximum singular value as T_j of the single-stream transmission user terminal; the second The executing module is configured to, for each dual-stream transmission user terminal whose transmission mode is determined to be dual-stream transmission by the first judging module, perform: determining H_j of the dual-stream transmission user terminal as T_j of the dual-stream transmission user terminal.

当该装置应用于FDD系统时，上述等效信道矩阵确定单元41具体可以划分为以下功能模块：When the device is applied to an FDD system, the above-mentioned equivalent channel matrix determination unit 41 can be specifically divided into the following functional modules:

码字索引获得模块，用于获得各个用户终端根据基站发送的下行公共导频信号CRS反馈的码字索引PMI；The codeword index obtaining module is used to obtain the codeword index PMI fed back by each user terminal according to the downlink common pilot signal CRS sent by the base station;

第一判断模块，用于分别判断针对各个用户终端的传输方式为单流传输或双流传输；The first judging module is used to respectively judge whether the transmission mode for each user terminal is single-stream transmission or dual-stream transmission;

第一执行模块，用于针对第一判断模块判断出传输方式为单流传输的每一单流传输用户终端，执行：将该单流传输用户终端反馈的PMI对应的预编码码字的第一列向量确定为该单流传输用户终端的T_j；The first execution module is configured to, for each single-stream transmission user terminal whose transmission mode is judged to be single-stream transmission by the first judging module, execute: the first The column vector is determined as T_j of the single-stream transmission user terminal;

第二执行模块，用于针对第一判断模块判断出传输方式为双流传输的每一双流传输用户终端，执行：根据该双流传输用户终端反馈的PMI对应的预编码码字的前两列向量确定该双流传输用户终端的T_j。The second execution module is configured to, for each dual-stream transmission user terminal whose transmission mode is determined to be dual-stream transmission by the first judgment module, execute: determine according to the first two column vectors of the precoding codeword corresponding to the PMI fed back by the dual-stream transmission user terminal The dual stream transmits T_j of the user terminal.

在一个较佳的实施例中，在用户终端的传输方式可能为单流传输也可能为双流传输的情况下，上述预编码向量确定单元46具体可以划分为以下功能模块：In a preferred embodiment, when the transmission mode of the user terminal may be single-stream transmission or dual-stream transmission, the above-mentioned precoding vector determination unit 46 may specifically be divided into the following functional modules:

第二判断模块，用于判断针对该用户终端的传输方式为单流传输或双流传输；确定模块，用于当第二判断模块判断出所述传输方式为单流传输时，将的最大奇异值对应的右奇异向量确定为预编码向量，当第二判断模块判断出所述传输方式为双流传输时，根据的最大和第二大奇异值分别对应的右奇异向量确定预编码向量。The second judgment module is used to judge whether the transmission method for the user terminal is single-stream transmission or dual-stream transmission; the determination module is used to determine that the transmission method is single-stream transmission when the second judgment module judges that the transmission method is single-stream transmission. The right singular vector corresponding to the largest singular value of is determined as the precoding vector, and when the second judging module judges that the transmission mode is dual-stream transmission, according to The right singular vectors corresponding to the largest and second largest singular values respectively determine the precoding vector.

对应于本发明实施例提供的一种应用于TDD系统的单用户终端传输方案中的预编码向量的确定方法，本发明实施例还提供一种基站，该基站的具体结构示意图如图5所示，包括以下功能单元：Corresponding to a method for determining a precoding vector applied to a single-user terminal transmission scheme of a TDD system provided by an embodiment of the present invention, an embodiment of the present invention also provides a base station, and a specific structural diagram of the base station is shown in FIG. 5 , including the following functional units:

矩阵与数量确定单元51，用于确定用户终端对应的下行信道矩阵H_i和用户终端接收天线的数量；A matrix and quantity determining unit 51, configured to determine the downlink channel matrix H_i corresponding to the user terminal and the number of receiving antennas of the user terminal;

简化SVD算法确定单元52，用于以矩阵与数量确定单元51确定的H_i作为待进行SVD的矩阵，从预先设置的接收天线数量与针对待进行SVD的矩阵执行简化SVD算法的对应关系中，确定矩阵与数量确定单元确定的该数量所对应的简化SVD算法；The simplified SVD algorithm determination unit 52 is used to use the H_i determined by the matrix and number determination unit 51 as the matrix to be SVD, from the corresponding relationship between the preset number of receiving antennas and the simplified SVD algorithm for the matrix to be performed SVD, Determine the simplified SVD algorithm corresponding to the quantity determined by the matrix and the quantity determination unit;

预编码向量确定单元53，用于根据简化SVD算法确定单元52确定的简化SVD算法确定H_i的奇异值对应的右奇异向量，确定针对所述用户终端的预编码向量。The precoding vector determining unit 53 is configured to determine the right singular vector corresponding to the singular value of H_i according to the simplified SVD algorithm determined by the simplified SVD algorithm determining unit 52, and determine the precoding vector for the user terminal.

在用户终端的传输方式可能为单流传输也可能为双流传输的情况下，上述预编码向量确定单元53具体可以划分为以下功能模块：In the case that the transmission mode of the user terminal may be single-stream transmission or dual-stream transmission, the above-mentioned precoding vector determination unit 53 may specifically be divided into the following functional modules:

判断模块，用于判断针对所述用户终端的传输方式为单流传输或双流传输；预编码向量确定模块，用于在判断模块判断出传输方式为单流传输时，将H_i的最大奇异值对应的右奇异向量确定为预编码向量，在判断模块判断出传输方式为双流传输时，根据H_i的最大和第二大奇异值分别对应的右奇异向量确定预编码向量。A judging module, configured to determine whether the transmission mode for the user terminal is single-stream transmission or dual-stream transmission; a precoding vector determination module, configured to set the maximum singular value of H_i when the judging module determines that the transmission mode is single-stream transmission The corresponding right singular vector is determined as the precoding vector, and when the judging module determines that the transmission mode is dual-stream transmission, the precoding vector is determined according to the right singular vectors corresponding to the largest and second largest singular values of H_i respectively.

对应于本发明实施例提供的一种应用于FDD系统的单用户终端传输方案中的预编码向量的确定方法，本发明实施例还提供一种用户终端，该用户终端的具体结构示意图如图6所示，包括以下功能单元：Corresponding to a method for determining a precoding vector applied to a single-user terminal transmission scheme of an FDD system provided by an embodiment of the present invention, an embodiment of the present invention also provides a user terminal, and a specific structural diagram of the user terminal is shown in Figure 6 As shown, it includes the following functional units:

矩阵与数量确定单元61，用于确定用户终端的下行信道矩阵H_i，并确定用户终端接收天线的数量；A matrix and quantity determining unit 61, configured to determine the downlink channel matrix H_i of the user terminal, and determine the number of receiving antennas of the user terminal;

简化SVD算法确定单元62，用于以矩阵与数量确定单元61确定的H_i作为待进行SVD的矩阵，从预先设置的接收天线数量与针对待进行SVD的矩阵执行简化SVD算法的对应关系中，确定与矩阵与数量确定单元61确定的该数量对应的简化SVD算法；The simplified SVD algorithm determination unit 62 is used to use the H_i determined by the matrix and quantity determination unit 61 as the matrix to be performed SVD, from the corresponding relationship between the preset number of receiving antennas and the simplified SVD algorithm for the matrix to be performed SVD, Determine the simplified SVD algorithm corresponding to the quantity determined by the matrix and quantity determination unit 61;

发送单元63，用于根据简化SVD算法确定单元62确定的简化SVD算法所确定的H_i的奇异值对应的右奇异向量，确定并向基站发送相应的PMI，指示基站根据PMI确定相应的预编码向量。The sending unit 63 is configured to determine and send the corresponding PMI to the base station according to the right singular vector corresponding to the singular value of_Hi determined by the simplified SVD algorithm determined by the simplified SVD algorithm determined by the simplified SVD algorithm determining unit 62, and instruct the base station to determine the corresponding precoding according to the PMI vector.

在一个较佳的实施例中，上述发送单元63具体可以划分为以下功能模块，包括：In a preferred embodiment, the above-mentioned sending unit 63 can specifically be divided into the following functional modules, including:

判断模块，用于判断针对用户终端的传输方式为单流传输或双流传输；A judging module, configured to judge whether the transmission mode for the user terminal is single-stream transmission or dual-stream transmission;

选取模块，用于当判断模块判断出传输方式为单流传输时，从预编码码本中，选取相应于H_i的最大奇异值所对应的右奇异向量的码字当断模块判断出传输方式为双流传输时，从预编码码本中选取相应于H_i的最大和第二大奇异值分所别对应的右奇异向量构成的右奇异矩阵的码字；The selection module is used to select the codeword corresponding to the right singular vector corresponding to the maximum singular value of H_i from the precoding codebook when the judging module judges that the transmission mode is single-stream transmission. When the judging module judges the transmission mode When it is a dual-stream transmission, select the codeword corresponding to the right singular matrix formed by the right singular vectors corresponding to the largest and second largest singular values of H_i from the precoding codebook;

确定模块，用于确定选取模块选取的码字对应的PMI；A determination module, configured to determine the PMI corresponding to the codeword selected by the selection module;

发送模块，用于向基站发送确定模块确定的PMI，指示基站根据发送模块发送的PMI确定相应的预编码向量。The sending module is configured to send the PMI determined by the determining module to the base station, and instruct the base station to determine a corresponding precoding vector according to the PMI sent by the sending module.

针对本发明实施例提供的上述装置、基站和用户终端，上述预先设置的接收天线数量与简化SVD算法的对应关系都可以但不限于为如下形式：For the above-mentioned device, base station, and user terminal provided in the embodiments of the present invention, the correspondence between the above-mentioned preset number of receiving antennas and the simplified SVD algorithm can be, but not limited to, the following form:

为1的接收天线数量对应的简化SVD算法为：以|H|作为对待进行SVD的矩阵H进行SVD后得到的最大奇异值，并以H|H|^-1作为对应于|H|的右奇异值向量，其中，|H|表示计算H的2-范数；The simplified SVD algorithm corresponding to the number of receiving antennas is 1: take |H| as the maximum singular value obtained after SVD on the matrix H to be subjected to SVD, and use H|H|^-1 as the right singular value corresponding to |H| Value vector, where |H| represents calculating the 2-norm of H;

为2的接收天线数量对应的简化SVD算法为：以λ₁²和分别作为对H进行SVD后得到的最大奇异值和第二大奇异值，并以V₁和V₂分别作为对应于λ₁²和的右奇异值向量，其中，λ₁²、V₁、V₂按照下述公式计算：The simplified SVD algorithm corresponding to the number of receiving antennas is 2: with λ₁² and Respectively as the largest singular value and the second largest singular value obtained after SVD on H, and take V₁ and V₂ as corresponding to λ₁² and The right singular value vector of , where, λ₁² , V₁ and V₂ are calculated according to the following formula:

${λ λ}_{11}^{22} = = \frac{{| | {h h}_{11} | |}^{22} {((β β + + \sqrt{11 - - {β β}^{22}} | | η η | |))}^{22} + + {| | {h h}_{22} | |}^{22} {((β β | | η η | | + + \sqrt{11 - - {β β}^{22}}))}^{22}}{11 + + 22 β β \sqrt{11 - - {β β}^{22}} | | η η | |}$

${λ λ}_{22}^{22} = = {| | {h h}_{11} | |}^{22} + + {| | {h h}_{22} | |}^{22} - - {λ λ}_{11}^{22}$

${V V}_{11} = = \frac{β β {h h}_{11} {| | {h h}_{11} | |}^{- - 11} + + \sqrt{11 - - {β β}^{22}} η η {| | η η | |}^{- - 11} {h h}_{22} {| | {h h}_{22} | |}^{- - 11}}{\sqrt{11 + + 22 β β \sqrt{11 - - {β β}^{22}} | | η η | |}}$

${V V}_{22} = = \frac{{h h}_{11} - - {h h}_{11} {V V}_{11}^{H h} {V V}_{11}}{| | {h h}_{11} - - {h h}_{11} {V V}_{11}^{H h} {V V}_{11} | |}$

其中，h₁为H的第一个行向量，h₂为H的第二个行向量，η和β按照下述公式计算：Among them, h₁ is the first row vector of H, h₂ is the second row vector of H, η and β are calculated according to the following formula:

$η η = = {h h}_{11} {h h}_{22}^{H h} {| | {h h}_{11} | |}^{- - 11} {| | {h h}_{22} | |}^{- - 11}$

$β β = = \sqrt{\frac{11}{22} \frac{22 {| | {h h}_{11} | |}^{22} (({| | {h h}_{11} | |}^{22} + + {| | {h h}_{22} | |}^{22})) {| | η η | |}^{22} + + {(({| | {h h}_{11} | |}^{22} - - {| | {h h}_{22} | |}^{22}))}^{22} + + (({| | {h h}_{11} | |}^{22} - - {| | {h h}_{22} | |}^{22})) \sqrt{{(({| | {h h}_{11} | |}^{22} - - {| | {h h}_{22} | |}^{22}))}^{22} + + 44 {| | {h h}_{11} | |}^{22} {| | {h h}_{22} | |}^{22} {| | η η | |}^{22}}}{{(({| | {h h}_{11} | |}^{22} - - {| | {h h}_{22} | |}^{22}))}^{22} + + {(({| | {h h}_{11} | |}^{22} + + {| | {h h}_{22} | |}^{22}))}^{22} {| | η η | |}^{22}}}$

；;

为4的接收天线数量对应的简化SVD算法为：The simplified SVD algorithm corresponding to the number of receiving antennas of 4 is:

以矩阵作为待进行SVD的矩阵，按照为2的接收天线数量对应的简化SVD算法，分别确定的最大奇异值、第二大奇异值、最大奇异值对应的右奇异向量、第二大奇异值对应的右奇异向量，作为H相应的最大奇异值、第二大奇异值、最大奇异值对应的右奇异向量、第二大奇异值对应的右奇异向量；其中，λ_1，1、V_1，1、λ_2，1、V_2，1按照下述方式确定：by matrix As the matrix to be SVD, according to the simplified SVD algorithm corresponding to the number of receiving antennas of 2, respectively determine The largest singular value, the second largest singular value, the right singular vector corresponding to the largest singular value, and the right singular vector corresponding to the second largest singular value, as the corresponding largest singular value, second largest singular value, and largest singular value of H The right singular vector and the right singular vector corresponding to the second largest singular value; among them, λ_1,1 , V_1,1 , λ_2,1 , V_2,1 are determined as follows:

对H进行划分得到H₁和H₂，其中H₁由H的两个行向量组成，H₂由H的另外两个行向量组成；以H₁和H₂作为待进行SVD的矩阵，按照为2的接收天线数量对应的简化SVD算法，分别确定H₁的最大奇异值λ_1，1与相应的右奇异向量V_1，1，以及H₂的最大奇异值λ_2，1与相应的右奇异向量V_2，1。Divide H to get H₁ and H₂ , where H₁ is composed of two row vectors of H, and H₂ is composed of the other two row vectors of H; H₁ and H₂ are used as the matrix to be SVD, according to The simplified SVD algorithm corresponding to the number of receiving antennas of 2 determines the maximum singular value λ_1,1 of H₁ and the corresponding right singular vector V_1,1 , and the maximum singular value λ_2,1 of H₂ and the corresponding right singular vector Vector V_2,1 .

显然，本领域的技术人员可以对本发明进行各种改动和变型而不脱离本发明的精神和范围。这样，倘若本发明的这些修改和变型属于本发明权利要求及其等同技术的范围之内，则本发明也意图包含这些改动和变型在内。Obviously, those skilled in the art can make various changes and modifications to the present invention without departing from the spirit and scope of the present invention. Thus, if these modifications and variations of the present invention fall within the scope of the claims of the present invention and equivalent technologies thereof, the present invention also intends to include these modifications and variations.