WO2014135075A1 - 检测方法及装置 - Google Patents

检测方法及装置 Download PDF

Info

Publication number
WO2014135075A1
WO2014135075A1 PCT/CN2014/072897 CN2014072897W WO2014135075A1 WO 2014135075 A1 WO2014135075 A1 WO 2014135075A1 CN 2014072897 W CN2014072897 W CN 2014072897W WO 2014135075 A1 WO2014135075 A1 WO 2014135075A1
Authority
WO
WIPO (PCT)
Prior art keywords
matrix
channel matrix
received signal
equivalent
vector
Prior art date
Application number
PCT/CN2014/072897
Other languages
English (en)
French (fr)
Inventor
余荣道
Original Assignee
华为技术有限公司
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 华为技术有限公司 filed Critical 华为技术有限公司
Publication of WO2014135075A1 publication Critical patent/WO2014135075A1/zh

Links

Classifications

    • HELECTRICITY
    • H04ELECTRIC COMMUNICATION TECHNIQUE
    • H04BTRANSMISSION
    • H04B7/00Radio transmission systems, i.e. using radiation field
    • H04B7/02Diversity systems; Multi-antenna system, i.e. transmission or reception using multiple antennas
    • H04B7/04Diversity systems; Multi-antenna system, i.e. transmission or reception using multiple antennas using two or more spaced independent antennas

Definitions

  • the present invention relates to the field of wireless communication technologies, and in particular, to a detection method and apparatus.
  • Mul t iple Input Mul t iple Output (MIMO) technology is a hotspot in the field of wireless communication.
  • MIM0 technology is used in various new mobile communication systems to improve the spectrum efficiency of the system.
  • MIM0 technology can increase the spatial dimension of data multiplexing, multiplex multiple copies of data space to the same time-frequency resource, or send the same data on multiple antennas and/or receive the same data with multiple receive antennas. Spatial diversity gain.
  • Typical space diversity techniques include
  • STBC Space Time Block Coding
  • typical space multiplexing technology includes Bell's vertical layered space-time technology (Vert ica l Bel l Labs Layered
  • V- BLAST V- BLAST
  • FIG. 1 is a schematic diagram of a MIMO detection application scenario.
  • a transmitting end transmits a transmitting signal through a transmitting world
  • a receiving end receives the signal through a receiving antenna, and detects a transmitting signal by using a MIMO technology, and the transmitting signal can use a transmitting signal.
  • Vector representation That is, the basic feature of the MIMO technology is multiple transmit antennas and multiple receive antennas. Assuming that the number of transmit antennas is ⁇ and the number of receive antennas M R , the MIMO transmission model can be expressed as:
  • the signal received on the first receiving antenna is the received signal vector composed of the received signal, /3 ⁇ 4. is the channel response between the first receiving antenna and the first transmitting antenna, and H is the channel
  • is the data symbol transmitted on the first transmitting antenna
  • s is the transmitted signal vector composed of the data symbols transmitted on the transmitting antenna, which is the noise received on the first receiving antenna, "is a noise matrix.
  • the transmitted signal vector s can be detected by MIMO detection. when When the number of receiving antennas is not less than the number of transmitted symbols, the receiving end can eliminate or suppress interference between multiple transmitted symbols by a certain MIMO equalization algorithm, thereby recovering a transmitted symbol.
  • the common linear MIMO equalization algorithm has a linear minimum.
  • LMMSE Linear Minimum Mean Square Error
  • ZF Zero Forcing
  • the receiver can also use all the transmitted symbols as a complete codeword using Maximum Likelihood Detection (MLD).
  • MMD Maximum Likelihood Detection
  • the method performs a test to estimate ⁇ ⁇ transmitted symbols.
  • MIMO equalization algorithm combined with Serial Interference Cancellation (SIC) for receiving, that is, first to estimate one of the transmitted symbols using a linear MIMO equalization method, and then use it as a known interference to eliminate the linear MIMO equalization method. Estimate another transmitted symbol and then iterate sequentially until all transmitted symbols are detected and received.
  • SIC Serial Interference Cancellation
  • ZF has the lowest complexity, but its performance is poor.
  • Embodiments of the present invention provide a detection method and apparatus, which can reduce detection complexity and improve performance.
  • an embodiment of the present invention provides a detection method, including:
  • Generating a first channel matrix by removing a set of column vectors from the original channel matrix, the original channel matrix corresponding to the received signal; calculating a first matrix corresponding to the first channel matrix, the first matrix including the first a singular vector corresponding to a singular value of a conjugate transposed matrix of a channel matrix;
  • the detecting, by using the equivalent received signal vector, the corresponding channel matrix, the transmit signal vector is: the equivalent received signal vector, Said equivalent channel matrix ⁇ detecting the transmitted signal by maximum likelihood detection MLD method Vector.
  • the conjugate transposed matrix of the first channel matrix is subjected to singular value decomposition to obtain the first matrix.
  • the removing a set of column vectors from the original channel matrix to generate the first channel matrix further includes: grouping the original channel matrix into columns, each column vector A group contains more than one column vector.
  • each of the column vector groups includes the same number of column vectors.
  • an embodiment of the present invention provides a detecting apparatus, including:
  • Generating unit which removes a set of column vectors from the original channel matrix to generate a first channel matrix, where the original channel matrix corresponds to the received signal;
  • a calculating unit configured to calculate a first matrix corresponding to the first channel matrix, where the first matrix includes a singular vector corresponding to a singular value of a conjugate transposed matrix of the first channel matrix;
  • a processing unit configured to perform multiplication processing on a conjugate transposed matrix of the first matrix and a received signal vector corresponding to the received signal to obtain an equivalent received signal vector, and a conjugate transposed matrix of the first matrix Performing multiplication processing with the original channel matrix to obtain an equivalent channel matrix;
  • a detecting unit configured to detect a transmit signal vector according to the equivalent received signal vector and the equivalent channel matrix.
  • the detecting unit is specifically configured to: detect the equivalent received signal vector, the equivalent channel matrix, and the maximum likelihood detection MLD method Transmit signal vector.
  • the calculating unit performs singular value decomposition on the conjugate transposed matrix of the first channel matrix to obtain the first matrix.
  • the generating unit is further configured to group the original channel matrix into columns, and each column vector group includes more than one column vector.
  • each column vector group contains the same number of column vectors.
  • the first channel matrix is generated by removing a set of column vectors from the original channel matrix, where the original channel matrix corresponds to the received signal; and the first matrix corresponding to the first channel matrix is calculated, a matrix comprising a singular vector corresponding to a singular value of a conjugate transposed matrix of the first channel matrix; multiplying a conjugate transposed matrix of the first matrix with a received signal vector corresponding to the received signal Obtaining an equivalent received signal vector, and multiplying the conjugate transposed matrix of the first matrix with the original channel matrix to obtain an equivalent channel matrix; according to the equivalent received signal vector, the equivalent channel matrix A transmitted signal vector is detected. Thereby, the detection complexity of the transmitted signal vector can be greatly reduced.
  • Figure 1 is a schematic diagram of an application scenario of MIM0 detection
  • FIG. 2 is a flowchart of a detection method according to Embodiment 1 of the present invention.
  • FIG. 3 is a comparison diagram of floating-point operands required by different algorithms according to Embodiment 1 of the present invention
  • FIG. 4 is a performance comparison diagram of different algorithms according to Embodiment 1 of the present invention
  • an embodiment of the present invention provides a detection method and apparatus.
  • the detection complexity can be reduced to the same order of magnitude as the ZF method. And the performance is much better than ZF.
  • Embodiment 1 of the present invention Method flow chart As shown in FIG. 2, the method includes:
  • a first channel matrix is generated by removing a set of column vectors from the original channel matrix, where the original channel matrix corresponds to the received signal.
  • a first channel matrix may be generated by removing a set of column vectors from the original channel matrix, the set of column vectors comprising more than one column vector.
  • the original channel matrix may also be grouped in columns before the step, and then one of the column vector groups is removed.
  • the number of column vectors contained in each column vector group can be the same or different.
  • the original channel matrix H can be represented by N column vector groups:
  • Each column vector group contains the one or more column vectors.
  • a first channel matrix may be generated after removing a set of column vectors from the original channel matrix, and the first channel matrix may be expressed as:
  • ⁇ ⁇ i ⁇ M T that is, means that the first column vector group is removed from H as the number of transmitting antennas, that is, the number of column vectors included in H, 1 ⁇
  • the number of receiving antennas that is, the number of row vectors included in H.
  • the first matrix composed of the vector corresponding to the singular value of the conjugate transposed matrix of the first channel matrix may be obtained by using the singular value decomposition method, or the first matrix may be obtained by other methods.
  • Singular value decomposition is a known method of matrix decomposition. It is not elaborated here. It is only briefly introduced: The singular value of the matrix can be decomposed into:
  • the matrix corresponding to the 0 singular value of ⁇ 1 is referred to herein as the first matrix. And according to the singular value decomposition characteristics in the linear theory, it can be known that:
  • multiplying the conjugate transposed matrix of the first matrix by the received signal vector is:
  • Multiplying the conjugate transposed matrix of the first matrix by the original channel matrix is:
  • the original channel matrix H corresponds to the received signal vector s. Therefore, when grouping the received signal vector s, the same grouping method as the original signal matrix should be used to make the grouped correspondence. The relationship remains the same.
  • equation (12) It can be seen from equation (12) that the solution to equation (1) can be transformed into the solution of the following N independent equations:
  • the transmission signal vector can be detected by the MLD method for the above equation group, and the MLD algorithm is:
  • the detected value of the transmitted signal vector s can be obtained.
  • the original MLD detection algorithm has a complexity of 2.8147e+014 at 64QAM and 8 transmit antennas. It can be seen that the complexity is greatly reduced.
  • FIG. 3 is a comparison diagram of floating point operands required by different algorithms according to Embodiment 1 of the present invention.
  • Figure 3 shows the simulation comparison diagram of the floating point operation (Flops) required by the scheme of the present application, the ZF method and the existing MLD detection algorithm. The more floating point numbers, the higher the computational complexity.
  • the simulation conditions of Fig. 3 are: 8 transmit antennas, 8 receive antennas, and the channel matrix H is divided into 1 column vector group.
  • "1" indicates the flops required for ZF, and its size is 1.6e05.
  • “2” indicates the flops required for the solution of the present invention, and its size is 2.2e05.
  • “3” indicates the flops required for the existing MLD method.
  • the size is 4.2e7. It can be seen that the solution of the present invention complicates the MIM0 detection algorithm. The degree is reduced to the same order of magnitude as ZF, and the performance is much better than ZF.
  • FIG. 4 is a performance comparison diagram of different algorithms according to Embodiment 1 of the present invention.
  • the simulation conditions are the same as those in Figure 3.
  • the vertical axis represents the bit error rate (BER) and the horizontal axis represents the signal-to-noise ratio (S igna l noisy s Ra t io , SNR ).
  • the same SNR corresponds.
  • the inventive scheme has a gain of nearly 5 dB with respect to ZF, and therefore, the performance is greatly improved compared to ZF.
  • a first channel matrix is generated by removing a set of column vectors from an original channel matrix, where the original channel matrix corresponds to a received signal; and calculating a first matrix corresponding to the first channel matrix,
  • the first matrix includes a singular vector corresponding to a 0 singular value of the conjugate transposed matrix of the first channel matrix; multiplying a conjugate transposed matrix of the first matrix by a received signal vector corresponding to the received signal Processing, obtaining an equivalent received signal vector, and multiplying the conjugate transposed matrix of the first matrix by the original channel matrix to obtain an equivalent channel matrix; according to the equivalent received signal vector, the equivalent channel
  • the matrix detects the transmitted signal vector. This greatly reduces the detection complexity of the transmitted signal vector and reduces the complexity to the same order of magnitude as the ZF method, and the performance is much better than ZF.
  • FIG. 5 is a schematic diagram of a detecting apparatus according to Embodiment 2 of the present invention. As shown in FIG. 5, the apparatus includes: a generating unit 501, a calculating unit 502, a processing unit 503, and a detecting unit 504.
  • the generating unit 501 is configured to generate a first channel matrix by removing a set of column vectors from the original channel matrix, where the original channel matrix corresponds to the received signal.
  • the calculating unit 502 is configured to calculate a first matrix corresponding to the first channel matrix, where the first matrix includes a singular vector corresponding to a singular value of a conjugate transposed matrix of the first channel matrix.
  • the processing unit 503 is configured to perform multiplication processing on the conjugate transposed matrix of the first matrix and the received signal vector corresponding to the received signal to obtain an equivalent received signal vector, and the conjugate transpose of the first matrix A matrix is multiplied with the original channel matrix to obtain an equivalent channel matrix.
  • a detecting unit 504 configured to detect, according to the equivalent received signal vector, the equivalent channel matrix The transmitted signal vector is measured.
  • the generating unit 501 is further configured to group the original channel matrix into columns, and each column vector group includes more than one column vector.
  • the detection unit 504 detects the complexity of the MLD device as N*2 M , where N is the number of column vector groups included in the original channel matrix, The number of column vectors included in the original channel matrix, and ⁇ is the number of constellation points used to modulate the transmitted signal corresponding to the transmitted signal vector.
  • the calculating unit 503 performs singular value decomposition on the conjugate transposed matrix of the first channel matrix to obtain the first matrix.
  • the detecting unit 504 is specifically configured to detect the transmit signal vector by using the maximum likelihood detection MLD method for the equivalent received signal vector and the equivalent channel matrix.
  • the detecting apparatus provided in the embodiment of the present invention is implanted with the detecting method provided in the first embodiment. Therefore, the specific working process of each unit in the detecting apparatus is not described herein.
  • a first channel matrix is generated by removing a set of column vectors from a source channel matrix by a generating unit, where the original channel matrix corresponds to a received signal; and the calculating unit calculates a first corresponding to the first channel matrix.
  • the first matrix includes a singular vector corresponding to a singular value of a conjugate transposed matrix of the first channel matrix; and the processing unit associates a conjugate transposed matrix of the first matrix with the received signal
  • the received signal vector is subjected to multiplication processing to obtain an equivalent received signal vector, and the conjugate transposed matrix of the first matrix is multiplied with the original channel matrix to obtain an equivalent channel matrix; the detecting unit receives the equivalent according to the equivalent A signal vector, the equivalent channel matrix, detects a transmitted signal vector. This greatly reduces the detection complexity of the transmitted signal vector and reduces the complexity to the same order of magnitude as the ZF method, and the performance is much better than ZF.
  • RAM random access memory
  • ROM read only memory
  • electrically programmable ROM electrically erasable programmable ROM
  • registers hard disk, removable disk, CD-ROM, or any other form of storage known in the art.

Landscapes

  • Engineering & Computer Science (AREA)
  • Computer Networks & Wireless Communication (AREA)
  • Signal Processing (AREA)
  • Radio Transmission System (AREA)

Abstract

本发明实施例涉及一种检测方法及装置,包括:通过从原始信道矩阵中去除一组列向量生成第一信道矩阵,所述原始信道矩阵与接收信号相对应;计算所述第一信道矩阵对应的第一矩阵,所述第一矩阵包括所述第一信道矩阵的共轭转置矩阵的0奇异值对应的奇异向量;将所述第一矩阵的共轭转置矩阵与所述接收信号对应的接收信号向量进行相乘处理获得等效接收信号向量,并且所述第一矩阵的共轭转置矩阵与所述原始信道矩阵进行相乘处理获得等效信道矩阵;根据所述等效接收信号向量、所述等效信道矩阵检测出发射信号向量。从而可以极大地降低发射信号向量的检测复杂度,并且可以将复杂度降低到与ZF方法同一个数量级,而且性能比ZF有很大改进。

Description

检测方法及装置
技术领域
本发明涉及无线通信技术领域, 尤其涉及一种检测方法及装置。
背景技术
多入多出 ( Mul t iple Input Mul t iple Output , MIMO )技术是无线通信 领域目前研究的热点, 各种新型移动通信系统中都釆用 MIM0技术来提高系 统的频谱效率。 MIM0技术可以增加数据复用的空间维度,使多份数据空间复 用到相同的时频资源, 也可以用多个天线上发送同样的数据和 /或用多个接 收天线接收同样的数据, 获得空间分集增益。 典型的空间分集技术包括
Alamout i空时分组码 ( Space Time Block Coding, STBC ), 而典型的空间复 用技术包括贝尔实险室垂直分层空时技术(Vert ica l Bel l Labs Layered
Space Time , V- BLAST )。
图 1为 MIMO检测应用场景示意图, 如图 1所示, 发射端将发射信号通 过发射天下发射出去, 接收端通过接收天线接收该信号, 并通过 MIMO技术 检测出发射信号, 发射信号可以用发射信号向量表示。 即 MIMO技术的基本 特征是多个发射天线和多个接收天线, 假设发射天线数为 Μ , 接收天线数 MR , 则可将 MIMO传输模型表示为:
Figure imgf000003_0001
或者简记为 y = + w。 其中 ,.为第 个接收天线上收到的信号,; 为由接 收到的信号组成的接收信号向量, /¾.为第 个接收天线与第 ·个发射天线之间 的信道响应, H为信道矩阵, ^为第 ·个发射天线上发送的数据符号, s为由 发射天线上发送的数据符号组成的发射信号向量, 为第 个接收天线上收 到的噪声, 《为噪声矩阵。 通过 MIMO检测可以检测出发射信号向量 s。 当 接收天线数不少于发射符号数时, 接收端能够通过一定的 MIMO 均衡算法 尽可能消除或抑制多个发射符号之间的干扰, 从而恢复出 个发射符号, 常见的线性 MIMO 均衡算法有线性最小均方误差 (Linear Minimum Mean Square Error, LMMSE )和迫零( Zero Forcing, ZF )等; 另外接收端也可以 将所有 个发射符号当成一个完整码字使用最大似然检测 (Maximum Likelihood Detection, MLD )方法进行检测, 从而估计出 Μτ个发射符号。 也 可以使用 MIMO 均衡算法结合串行干扰消除 ( Successive Interference Cancellation, SIC )进行接收, 即首先使用线性 MIMO均衡方法估计其中一 个发射符号, 然后将其作为已知干扰进行消除后再使用线性 MIMO 均衡方 法估计另外一个发射符号, 然后依次迭代, 直到所有发射符号都检测接收完 毕。
在 MIMO的所有检测算法中, ZF复杂度最低, 但其性能较差。
发明内容
本发明实施例提供了一种检测方法及装置, 既可以降低检测复杂度, 又 可以提高性能。
在第一方面, 本发明实施例提供了一种检测方法, 包括:
从原始信道矩阵中去除一组列向量生成第一信道矩阵,所述原始信道矩 阵与接收信号相对应; 计算所述第一信道矩阵对应的第一矩阵,所述第一矩阵包括所述第一信 道矩阵的共轭转置矩阵的 0奇异值对应的奇异向量;
将所述第一矩阵的共轭转置矩阵与所述接收信号对应的接收信号向量 进行相乘处理获得等效接收信号向量, 并且所述第一矩阵的共轭转置矩阵与 所述原始信道矩阵进行相乘处理获得等效信道矩阵;
根据所述等效接收信号向量、 所述等效信道矩阵检测出发射信号向量。 结合第一方面, 在第一种可能的实现方式中, 所述根据所述等效接收信 号向量、 所述等效信道矩阵检测出发射信号向量具体为: 对所述等效接收信 号向量、所述等效信道矩阵釆用最大似然检测 MLD方法检测出所述发射信号 向量。
结合第一方面, 在第二种可能的实现方式中,对所述第一信道矩阵的共 轭转置矩阵进行奇异值分解, 获得所述第一矩阵。
结合第一方面, 在第三种可能的实现方式中, 所述从原始信道矩阵中去 除一组列向量生成第一信道矩阵之前还包括: 将所述原始信道矩阵按列分 组, 每个列向量组包含一个以上的列向量。
结合第三种可能的实现方式, 在第四种可能的实现方式中, 所述每个列 向量组包含相同数量的列向量。
在第二方面, 本发明实施例提供了一种检测装置, 包括:
生成单元,从原始信道矩阵中去除一组列向量生成第一信道矩阵, 所述 原始信道矩阵与接收信号相对应;
计算单元, 用于计算所述第一信道矩阵对应的第一矩阵, 所述第一矩阵 包括所述第一信道矩阵的共轭转置矩阵的 0奇异值对应的奇异向量;
处理单元,用于将所述第一矩阵的共轭转置矩阵与所述接收信号对应的 接收信号向量进行相乘处理获得等效接收信号向量, 并且所述第一矩阵的共 轭转置矩阵与所述原始信道矩阵进行相乘处理获得等效信道矩阵;
检测单元, 用于根据所述等效接收信号向量、 所述等效信道矩阵检测出 发射信号向量。
结合第二方面, 在第一种可能的实现方式中, 所述检测单元具体用于, 对所述等效接收信号向量、所述等效信道矩阵釆用最大似然检测 MLD方法检 测出所述发射信号向量。
结合第二方面, 在第二种可能的实现方式中, 所述计算单元对所述第一 信道矩阵的共轭转置矩阵进行奇异值分解, 获得所述第一矩阵。
结合第二方面, 在第三种可能的实现方式中, 所述生成单元还用, 将所 述原始信道矩阵按列分组, 每个列向量组包含一个以上的列向量。
结合第三种可能的实现方式, 在第四种可能的实现方式中, 所述生成单 元中, 所述每个列向量组包含相同数量的列向量。 本发明实施例中,通过从原始信道矩阵中去除一组列向量生成第一信道 矩阵, 所述原始信道矩阵与接收信号相对应; 计算所述第一信道矩阵对应的 第一矩阵, 所述第一矩阵包括所述第一信道矩阵的共轭转置矩阵的 0奇异值 对应的奇异向量; 将所述第一矩阵的共轭转置矩阵与所述接收信号对应的接 收信号向量进行相乘处理获得等效接收信号向量, 并且所述第一矩阵的共轭 转置矩阵与所述原始信道矩阵进行相乘处理获得等效信道矩阵; 根据所述等 效接收信号向量、 所述等效信道矩阵检测出发射信号向量。 从而可以极大地 降低发射信号向量的检测复杂度。
附图说明
为了更清楚地说明本发明实施例或现有技术中的技术方案, 下面将对实 施例或现有技术描述中所需要使用的附图作简单地介绍, 显而易见地, 对于 本领域普通技术人员而言, 在不付出创造性劳动性的前提下, 还可以根据这 些附图获得其他的附图。
图 1为 MIM0检测应用场景示意图;
图 2为本发明实施例一提供的检测方法流程图;
图 3为本发明实施例一提供的不同算法所需的浮点操作数对比图; 图 4为本发明实施例一提供的不同算法的性能对比图; 图 5为本发明实施例二提供的检测装置示意图。
具体实施方式
为使本发明的目的、 技术方案和优点更加清楚, 下面结合附图对本发明 具体实施例作进一步的详细描述。
针对现有技术的缺陷, 本发明实施例提出一种检测方法及装置, 通过将 原始信道矩阵分组, 在零空间进行求解检测出发射信号向量, 可以将检测复 杂度降低到与 ZF方法同一个数量级, 而且性能比 ZF有很大改进。
下述实施例描述的为一种检测方法。图 2为本发明实施例一提供的检测 方法流程图。 如图 2所示, 所述方法包括:
S201 ,从原始信道矩阵中去除一组列向量生成第一信道矩阵, 所述原始 信道矩阵与接收信号相对应。
具体地, 可以从原始信道矩阵中去除一组列向量生成第一信道矩阵, 该 一组列向量包括一个以上的列向量。
为操作方便, 在该步骤之前还可以先将所述原始信道矩阵按列分组, 然 后去掉其中的一个列向量组。每个列向量组包含的列向量数量可以相同也可 以不同。 具体为, 根据公式(1 )所示的 MIM0传输模型, 可以将原始信道矩 阵 H用 N个列向量组来表示:
H
Figure imgf000007_0001
每个列向量组包含所述一个以上的列向量。从原始信道矩阵中去除一组列向 量后可以生成第一信道矩阵, 所述第一信道矩阵可以表示为:
A H ΗΜ... ΗΝ] (3) 其中, \≤i≤MT , 即 .表示从 H去掉了第 个列向量组 为发射天线数, 即 H所包含的列向量数量, 1^为接收天线数, 即 H所包含的行向量数量。
S202 , 计算所述第一信道矩阵对应的第一矩阵, 所述第一矩阵包括所述 第一信道矩阵的共轭转置矩阵的 0奇异值对应的奇异向量。
具体地,可以釆用奇异值分解的方法求出第一信道矩阵 的共轭转置矩 阵 的 0奇异值对应的向量所组成的第一矩阵, 也可以釆用其他方法获得 第一矩阵。 奇异值分解是一种已知的矩阵分解方法, 在此不详细阐述, 只做 简要介绍: 可以将矩阵 进行奇异值分解为:
HH=
Figure imgf000007_0002
其中 ^ 1为 的 0奇异值对应的矩阵, 这里称为第一矩阵。 并且根据线性理 论中的奇异值分解特点, 可知:
Figure imgf000008_0001
为表述方便, 可令 = °, 则有
Figure imgf000008_0002
将公式( 6 )取共轭可得:
W^Hl =0 (7)
5203,将所述第一矩阵的共轭转置矩阵与所述接收信号对应的接收信号 向量进行相乘处理获得等效接收信号向量, 并且所述第一矩阵的共轭转置矩 阵与所述原始信道矩阵进行相乘处理获得等效信道矩阵。
具体地, 将所述第一矩阵的共轭转置矩阵与接收信号向量相乘为:
= η (8) 其中, ; 为接收信号对应的接收信号向量, 为等效接收信号向量。
将第一矩阵的共轭转置矩阵与原始信道矩阵相乘为:
W^H^H^ HM...HN]
Figure imgf000008_0003
其中, ^=^ H,., 称为等效信道矩阵。
5204,根据所述等效接收信号向量、 所述等效信道矩阵检测出发射信号 向量。 将如公式( 1 )所表示的 MIM0传输模型两边同时乘以上述矩阵 H , 可 得:
(10)
= Wi H([H ..Hi_l ^ HM...HN]s + n) 其中, 将接收信号向量 s按照与原始信道矩阵 H一样的分组方法, 将其分为 如下形式: 5 = [^.. .... ]τ, 为行向量组, 每个行向量组 包含的元素个数 与原始信道矩阵 Η中的每个列向量组 包含的元素个数相同。
需要说明的是, 原始信道矩阵 H与接收信号向量 s是相对应的, 因此, 这里对接收信号向量 s分组时, 也应该釆用与对原始信号矩阵一样的分组方 法, 使其分组后的对应关系保持不变。
则将公式( 8 )和公式( 9 ) 带入公式( 10 ) 中可得:
其中, = f^«, 为等效噪声向量。
从公式(12) 可以看出, 对公式(1 ) 的求解, 可以转化为对下面 N个 独立的方程的求解:
η =^.^+«, z=l,2,...,N (12) 优选地, 可以对上述方程组釆用 MLD方法检测出所述发射信号向量, MLD算法为:
St =argmin(^.-H;-S; ') , i = \,2 .,N (13) 其中, 上述公式求出的 为使 -^· 『取最小值时的 , 为 的估计 值, 或者说检测值, Ω为星座空间的星座点集合, 不同的星座调制方式, 其 大小不一样, 比如釆用正交相移键控 (Quadrature Phase Shift Keying, QPSK)调制, 有 4个星座点, 16相正交振幅调制 (Quadrature Amplitude Modulation, QAM)有 16个星座点, 64QAM有 64个星座点。 MLD检测算法的 复杂度公式为 Q。 , ρ为对所述发射信号向量对应的发射信号进行调制所用的 星座点个数, α为 所包含的元素个数。
通过公式(13)进行 N此检测后, 可以获得发射信号向量 s的检测值 当对原始信道矩阵 H分组时, 如果每个列向量组包含相等数量的列向 量, 釆用 MLD方法检测时, 由于^ =^ H,., 因此, 矩阵^的规模为
Ν Ν
(行数 X列数), 而 中一个行向量包含的元素个数与 A.的列向量包含的元 素个数是相等的, 由此可知 A中元素个数为 ^, 因此, 对 A检测的复杂度为
N
QM , 由于有 N个方程, 因此, 总的复杂度为 N*2M , N为所述原始信道 矩阵包含的列向量组数量。如果釆用现有的 MLD方法进行检测,则将公式( 1 ) 带入检测公式为: s = argmin( r -H -s ) (14) 可以看出, 检测复杂度为 ρ^, 因此, 本发明实施例提供的方法相对于 原有的 MLD检测算法来说, 复杂度得到了极大的降低。
具体地,对于本发明实施例提供的检测方法,如果釆用 64QAM星座调制, 8 发射天线, 将原始信道矩阵 H分块分为 1 个列块时, 复杂度为 2*64Λ (8/2)=3.3554432e+07,如果釆用 64QAM星座调制, 8发射天线,将原始 信道矩阵 H分块分为 2个列块时, 复杂度为 4*64Λ(8/4)=1.6384e+04, 而原 有的 MLD检测算法在 64QAM, 8发射天线时复杂度为 2.8147e+014,可以看出, 复杂度大大降低。
需要说明的是, 也可以釆用其它方法求解上述方程组, 来检测获得发射 信号向量 ^
图 3为本发明实施例一提供的不同算法所需的浮点操作数对比图。 图 3 所示的为本申请的方案、 ZF 方法和现有 MLD检测算法所需的浮点操作数 (Floating point operation, Flops)的仿真对比图, 浮点数越多说明计算 复杂度越高。 图 3的仿真条件为: 8发射天线, 8接收天线, 将信道矩阵 H分 为 1个列向量组。 图 3中 "1"表示 ZF所需的 flops, 其大小为 1.6e05, "2" 表示本发明方案所需的 flops, 其大小为 2.2e05, "3" 表示现有 MLD方法所 需的 flops其大小为 4.2e7。 由此可知,本发明的方案将 MIM0检测算法复杂 度降低到了与 ZF同一数量级, 而且性能比 ZF有很大改进。
图 4为本发明实施例一提供的不同算法的性能对比图。 仿真条件与图 3 中的条件相同, 纵轴表示比特误码率 (Bi t error ra t io, BER ) ,横轴表示 信噪比 (S igna l Noi se Ra t io , SNR ) , 同一 SNR对应的 BER越大说明性能 越差, 从中可以看出, 本发明方案相对于 ZF来说有将近 5dB的增益, 因此, 性能较之 ZF有很大改进。
上述实施例描述的为,通过从原始信道矩阵中去除一组列向量生成第一 信道矩阵, 所述原始信道矩阵与接收信号相对应; 计算所述第一信道矩阵对 应的第一矩阵, 所述第一矩阵包括所述第一信道矩阵的共轭转置矩阵的 0奇 异值对应的奇异向量; 将所述第一矩阵的共轭转置矩阵与所述接收信号对应 的接收信号向量进行相乘处理获得等效接收信号向量, 并且所述第一矩阵的 共轭转置矩阵与所述原始信道矩阵进行相乘处理获得等效信道矩阵; 根据所 述等效接收信号向量、 所述等效信道矩阵检测出发射信号向量。 从而可以极 大地降低发射信号向量的检测复杂度, 并且可以将复杂度降低到与 ZF方法 同一个数量级, 而且性能比 ZF有很大改进。
相应地, 本发明实施例提供了一种与上述检测方法对应的检测装置。 图 5为本发明实施例二提供的检测装置示意图。 如图 5所示, 所述装置包括: 生成单元 501、 计算单元 502、 处理单元 503和检测单元 504。
生成单元 501 , 用于从原始信道矩阵中去除一组列向量生成第一信道矩 阵, 所述原始信道矩阵与接收信号相对应。
计算单元 502 , 用于计算所述第一信道矩阵对应的第一矩阵, 所述第一 矩阵包括所述第一信道矩阵的共轭转置矩阵的 0奇异值对应的奇异向量。
处理单元 503 , 用于将所述第一矩阵的共轭转置矩阵与所述接收信号对 应的接收信号向量进行相乘处理获得等效接收信号向量, 并且所述第一矩阵 的共轭转置矩阵与所述原始信道矩阵进行相乘处理获得等效信道矩阵。
检测单元 504 , 用于根据所述等效接收信号向量、 所述等效信道矩阵检 测出发射信号向量。
其中, 所述生成单元 501还用于, 将所述原始信道矩阵按列分组, 每个 列向量组包含一个以上的列向量。
如果每个列向量组包含相同数量的列向量, 则所述检测单元 504 釆用 MLD 装置检测的复杂度为 N*2M , N为所述原始信道矩阵包含的列向量组 数量, 为所述原始信道矩阵包含的列向量个数, ρ为对所述发射信号向量 对应的发射信号进行调制所用的星座点个数。
所述计算单元 503 对所述第一信道矩阵的共轭转置矩阵进行奇异值分 解, 获得所述第一矩阵。
所述检测单元 504具体用于, 对所述等效接收信号向量、 所述等效信道 矩阵釆用最大似然检测 MLD方法检测出所述发射信号向量。
需要说明的是,本发明实施例提供的检测装置植入了实施例一提供的检 测方法, 因此, 所述检测装置中各个单元的具体工作过程在此不再赘述。
上述实施例描述的为,通过生成单元从原始信道矩阵中去除一组列向量 生成第一信道矩阵, 所述原始信道矩阵与接收信号相对应; 计算单元计算所 述第一信道矩阵对应的第一矩阵, 所述第一矩阵包括所述第一信道矩阵的共 轭转置矩阵的 0奇异值对应的奇异向量; 处理单元将所述第一矩阵的共轭转 置矩阵与所述接收信号对应的接收信号向量进行相乘处理获得等效接收信 号向量, 并且所述第一矩阵的共轭转置矩阵与所述原始信道矩阵进行相乘处 理获得等效信道矩阵; 检测单元根据所述等效接收信号向量、 所述等效信道 矩阵检测出发射信号向量。 从而可以极大地降低发射信号向量的检测复杂 度, 并且可以将复杂度降低到与 ZF方法同一个数量级, 而且性能比 ZF有很 大改进。
专业人员应该还可以进一步意识到,结合本文中所公开的实施例描述的 各示例的单元及算法步骤, 能够以电子硬件、 计算机软件或者二者的结合来 实现, 为了清楚地说明硬件和软件的可互换性, 在上述说明中已经按照功能 ^ ^ 一般性地描述了各示例的组成及步骤。 这些功能究竟以硬件还是软件方式来 执行, 取决于技术方案的特定应用和设计约束条件。 专业技术人员可以对每 个特定的应用来使用不同方法来实现所描述的功能,但是这种实现不应认为 超出本发明的范围。
结合本文中所公开的实施例描述的方法或算法的步骤可以用硬件、处理 器执行的软件模块, 或者二者的结合来实施。 软件模块可以置于随机存储器
( RAM )、 内存、 只读存储器(ROM )、 电可编程 R0M、 电可擦除可编程 R0M、 寄存器、 硬盘、 可移动磁盘、 CD-R0M、 或技术领域内所公知的任意其它形式 的存储介质中。 以上所述的具体实施方式, 对本发明的目的、 技术方案和有益效果进行 了进一步详细说明, 所应理解的是, 以上所述仅为本发明的具体实施方式而 已, 并不用于限定本发明的保护范围, 凡在本发明的精神和原则之内, 所做 的任何修改、 等同替换、 改进等, 均应包含在本发明的保护范围之内。

Claims

权利要求
1. 一种检测方法, 其特征在于, 所述方法包括:
从原始信道矩阵中去除一组列向量生成第一信道矩阵,所述原始信道矩 阵与接收信号相对应;
计算所述第一信道矩阵对应的第一矩阵,所述第一矩阵包括所述第一信 道矩阵的共轭转置矩阵的 0奇异值对应的奇异向量;
将所述第一矩阵的共轭转置矩阵与所述接收信号对应的接收信号向量 进行相乘处理获得等效接收信号向量, 并且所述第一矩阵的共轭转置矩阵与 所述原始信道矩阵进行相乘处理获得等效信道矩阵;
根据所述等效接收信号向量、 所述等效信道矩阵检测出发射信号向量。
2. 根据权利要求 1所述的方法, 其特征在于, 所述根据所述等效接收 信号向量、 所述等效信道矩阵检测出发射信号向量具体为: 对所述等效接收 信号向量、所述等效信道矩阵釆用最大似然检测 MLD方法检测出所述发射信 号向量。
3. 根据权利要求 1所述的方法, 其特征在于, 对所述第一信道矩阵的 共轭转置矩阵进行奇异值分解, 获得所述第一矩阵。
4. 根据权利要求 1所述的方法, 其特征在于, 所述从原始信道矩阵中 去除一组列向量生成第一信道矩阵之前还包括: 将所述原始信道矩阵按列分 组, 每个列向量组包含一个以上的列向量。
5. 根据权利要求 4所述的方法, 其特征在于, 所述每个列向量组包含 相同数量的列向量。
6. 一种检测装置, 其特征在于, 所述装置包括:
生成单元,从原始信道矩阵中去除一组列向量生成第一信道矩阵, 所述 原始信道矩阵与接收信号相对应;
计算单元, 用于计算所述第一信道矩阵对应的第一矩阵, 所述第一矩阵 包括所述第一信道矩阵的共轭转置矩阵的 0奇异值对应的奇异向量; 处理单元,用于将所述第一矩阵的共轭转置矩阵与所述接收信号对应的 接收信号向量进行相乘处理获得等效接收信号向量, 并且所述第一矩阵的共 轭转置矩阵与所述原始信道矩阵进行相乘处理获得等效信道矩阵;
检测单元, 用于根据所述等效接收信号向量、 所述等效信道矩阵检测出 发射信号向量。
7. 根据权利要求 6所述的装置,其特征在于,所述检测单元具体用于, 对所述等效接收信号向量、所述等效信道矩阵釆用最大似然检测 MLD装置检 测出所述发射信号向量。
8. 根据权利要求 6所述的装置, 其特征在于, 所述计算单元对所述第 一信道矩阵的共轭转置矩阵进行奇异值分解, 获得所述第一矩阵。
9. 根据权利要求 6所述的装置, 其特征在于, 所述生成单元还用, 将 所述原始信道矩阵按列分组, 每个列向量组包含一个以上的列向量。
10. 根据权利要求 9所述的装置, 其特征在于, 所述生成单元中, 所 述每个列向量组包含相同数量的列向量。
PCT/CN2014/072897 2013-03-05 2014-03-05 检测方法及装置 WO2014135075A1 (zh)

Applications Claiming Priority (2)

Application Number Priority Date Filing Date Title
CN201310069939.9A CN104038265B (zh) 2013-03-05 2013-03-05 检测方法及装置
CN201310069939.9 2013-03-05

Publications (1)

Publication Number Publication Date
WO2014135075A1 true WO2014135075A1 (zh) 2014-09-12

Family

ID=51468874

Family Applications (1)

Application Number Title Priority Date Filing Date
PCT/CN2014/072897 WO2014135075A1 (zh) 2013-03-05 2014-03-05 检测方法及装置

Country Status (2)

Country Link
CN (1) CN104038265B (zh)
WO (1) WO2014135075A1 (zh)

Families Citing this family (1)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN111669206B (zh) * 2020-05-21 2021-04-06 中国联合网络通信集团有限公司 一种信号检测方法及装置

Citations (5)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN102035624A (zh) * 2009-09-30 2011-04-27 华为技术有限公司 一种基于双时空发射分集系统的预编码方法及装置
CN102195757A (zh) * 2010-03-19 2011-09-21 华为技术有限公司 分布式多天线系统中的预编码、解码方法和装置
CN102647259A (zh) * 2012-04-07 2012-08-22 西安电子科技大学 基于奇异值分解的多用户mimo系统上行链路传输方法
CN102868433A (zh) * 2012-09-10 2013-01-09 西安电子科技大学 多输入多输出y信道中基于天线选择的信号传输方法
CN102882655A (zh) * 2012-10-29 2013-01-16 北京邮电大学 多天线系统基于网络编码的基站与用户联合传输方法

Family Cites Families (4)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
KR101015727B1 (ko) * 2007-07-20 2011-02-22 삼성전자주식회사 무선 통신 시스템에서 최대우도 검출 장치 및 방법
CN101127548A (zh) * 2007-08-18 2008-02-20 姜永权 非协作虚拟mimo技术
CN101888287B (zh) * 2009-05-14 2014-12-03 株式会社Ntt都科摩 一种多输入多输出接收机信号检测方法及装置
CN104184505B (zh) * 2013-05-23 2017-10-10 华为技术有限公司 发射信号的多输入多输出mimo检测方法、装置及系统

Patent Citations (5)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN102035624A (zh) * 2009-09-30 2011-04-27 华为技术有限公司 一种基于双时空发射分集系统的预编码方法及装置
CN102195757A (zh) * 2010-03-19 2011-09-21 华为技术有限公司 分布式多天线系统中的预编码、解码方法和装置
CN102647259A (zh) * 2012-04-07 2012-08-22 西安电子科技大学 基于奇异值分解的多用户mimo系统上行链路传输方法
CN102868433A (zh) * 2012-09-10 2013-01-09 西安电子科技大学 多输入多输出y信道中基于天线选择的信号传输方法
CN102882655A (zh) * 2012-10-29 2013-01-16 北京邮电大学 多天线系统基于网络编码的基站与用户联合传输方法

Also Published As

Publication number Publication date
CN104038265A (zh) 2014-09-10
CN104038265B (zh) 2017-11-21

Similar Documents

Publication Publication Date Title
TWI810373B (zh) 接收器、其操作方法、其訊號檢測方法和非暫時性電腦可讀記錄媒體
US9191080B2 (en) Reception device, transmission device, reception method, transmission method, program, and radio communication system
JP5053378B2 (ja) 等化構造及び等化方法
Hu et al. Low-complexity signal detection using CG method for uplink large-scale MIMO systems
CN107624235B (zh) 用于估计无线通信系统中的下行链路信道的装置和方法
TW201503623A (zh) 非線性預編碼為基礎之多使用者多輸入多輸出方法及程序
TWI591973B (zh) A signal detection method and device
JP5053377B2 (ja) 等化構造及び等化方法
JP2005328312A (ja) チャネル推定装置、チャネル推定方法及び無線受信機
JP2009152877A (ja) 無線通信システム、受信装置、受信方法
JP2020010406A (ja) スパースオーダーリングされた反復グループ判定帰還干渉除去
JP2008205697A (ja) Mimo受信装置および受信方法
JP6063607B2 (ja) 送信信号のための多入力多出力mimo検出方法、装置及びシステム
JP5859913B2 (ja) 無線受信装置、無線送信装置、無線通信システム、プログラムおよび集積回路
Hussein et al. Least Square Estimation‐Based Different Fast Fading Channel Models in MIMO‐OFDM Systems
CN112887068B (zh) 数据传输方法、发送设备和接收设备
JP2008283393A (ja) Mmse等化回路、受信装置、及び通信システム、並びにその方法及びプログラム
JP5770558B2 (ja) 受信装置、プログラムおよび集積回路
JP5047289B2 (ja) 等化構造及び等化方法
WO2017116292A1 (en) Method and receiving node for detecting signals transmitted by multiple users
CN108418619B (zh) 一种信号检测方法及装置
WO2014135075A1 (zh) 检测方法及装置
KR101238919B1 (ko) 공간 다중화 다중안테나 시스템에서의 신호 대 간섭 및 잡음비 예측 장치 및 방법
CN113315551B (zh) 分层空时码系统的信号检测方法、装置和计算机设备
JP6120595B2 (ja) Mimo受信装置及びプログラム

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: 14759881

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: 14759881

Country of ref document: EP

Kind code of ref document: A1