CN102594425B - Minimum-distance pre-coding method suitable for double-flow data transmission - Google Patents

Abstract

The invention discloses a minimum-distance pre-coding method suitable for double-flow data transmission. The minimum-distance pre-coding method is characterized in that the condition number of channel matrixes and the modulation order of quadrature amplitude modulation signals are used as index parameters, design parameters are obtained through looking up a pre-constructed pre-coding parameter look-up table suitable for quadrature amplitude modulation signals (4-256 orders), and an operation of carrying out real-time design of a minimum-distance pre-coding matrix is performed according to a closure calculation formula of rotating angles and power dividing angle parameters. Compared with the existing relevant design methods based on numerical exhaustive search strategy, the method disclosed by the invention not only has an approximately optimal pre-coding performance, but also has extremely low online computational complexity and engineering implementation costs, thereby facilitating the real-time implementation of the method in new-generation broadband wireless and mobile communication systems such as 802.11n, TD-HSPA+, TD-LTE and TD-LTE- Advanced and the like.

Description

A kind of minimum range method for precoding of applicable double-flow data transmission

Technical field

The invention belongs to multiple-input and multiple-output (MIMO) broadband wireless and mobile communication technology field, be specifically related to be applicable to the minimum range method for precoding of the applicable double-flow data transmission of the new generation broadband wireless such as 802.11n, TD-HSPA+, TD-LTE and TD-LTE-Advanced and mobile communication system.

Background technology

The in the situation that of transmitting terminal and receiving terminal known channel state information, MIMO precoding technique is spectrum efficiency and the link reliability of elevator system effectively.If receiving terminal adopts non-linear maximum likelihood receiver, the MIMO method for precoding based on minimax Euclid distance criterion is the effective measures of improving receiving-transmitting chain transmission quality, has caused industry extensive concern.From problem model, due to the nondeterministic polynomial difficult problem (NP-Hard) normally of the minimum range precoding optimization problem based on transmitted power constraint, even if develop high-performance solution in lower dimensional space, also has technological challenge.International IEEE-the signal of < < is processed transactions > > (IEEE Transactions on Signal Processing, vol.52, no.3, pp.617 – 627, Mar.2004) in disclosed " the optimum minimum range method for precoding that is applicable to MIMO SDM system " (Optimal minimum distance-based precoder for MIMO spatial multiplexing systems) literary composition, utilize first three angle parameters to characterize physical channel matrix and pre-coding matrix, a kind of method for precoding based on numerical value exhaustive search proposing can obtain the complex value pre-coding matrix that is applicable to double-flow data transmission.Yet the method, owing to being limited by adopted parameterized model and computation complexity, can only be applicable to the low-order digit modulation systems such as two-phase PSK (BPSK) and quarternary phase-shift keying (QPSK) (QPSK).In recent years, many researchers still expand the method in succession based on numerical value exhaustive search strategy, to adapt to more high dimensional data flow transmission and more high-order digit modulation system, wherein the international IEEE-signal of < < is processed transactions > > (IEEE Transactions on Signal Processing, vol.59, no.11, pp.5485 – 5498, Nov.2011) disclosed " a kind of low maximum-likelihood decoding complexity, full diversity gain full rate MIMO method for precoding " (A low ML-decoding complexity, full diversity, full-rate MIMO precoder) due to former complex value pre-coding matrix is defined as to real matrix, the method for precoding that proposed thus can reduce implementation complexity to a certain extent, but still need carry out numerical search for two angle parameters, real-time is poor.

In sum, existing minimum range method for precoding all carrys out computation model parameter based on numerical value exhaustive search strategy, lack corresponding closed expression formula, thereby its computation complexity is subject to the restrictions such as mimo channel and signal space dimension, numerical search precision and step-size in search, engineering using value is lower, cannot in new generation broadband wireless and mobile communication system, effectively implement, urgently develop the low complex degree minimum range MIMO method for precoding that real-time is higher.

Summary of the invention

The present invention proposes a kind of minimum range method for precoding of applicable double-flow data transmission, and to improve, the complexity existing in existing minimum range method for precoding is high, real-time is poor, is difficult to the problem that through engineering approaches is applied to practical communication system.

The present invention is applicable to the minimum range method for precoding of double-flow data transmission, establishes transmitting terminal and adopts quadrature amplitude modulation (QAM) to send two paths of data stream, and total transmitted power is P0, the equivalent virtual channel matrix of multiple-input and multiple-output (MIMO) flat fading channel and known at transmitting terminal, pre-coding matrix it is characterized in that concrete operation step is:

The first step, calculates channel matrix conditional number step: utilize singular value decomposition definition H=U Λ V ^*equivalent virtual channel matrix H is carried out to singular value decomposition, and wherein, left singular matrix U and the right singular matrix V of equivalent virtual channel matrix H are unitary matrix, the maximum singular value λ that comprises descending arrangement in the diagonal matrix Λ of equivalent virtual channel matrix H ₁with inferior large singular value λ ₂, V ^*expression is carried out conjugate transpose operation to matrix V; According to matrix conditional number definition κ=λ ₁/ λ ₂calculate channel matrix conditional number κ;

Second step, parameter finding step: utilize the Precoding Design parameter look-up table table 1 of applicable quadrature amplitude modulation given below (QAM) signal,

Table 1 is applicable to the Precoding Design parameter look-up table of quadrature amplitude modulation (QAM) signal

Rotation angle θ, the first auxiliary ginseng a, the second auxiliary ginseng b and the concrete numerical value of the 3rd auxiliary ginseng c are determined in the interval falling into according to order of modulation M and channel matrix conditional number κ, change text description into:

When order of modulation M=4, if channel matrix conditional number κ ∈ [1,2.6455], rotation angle θ=π/4, the first auxiliary ginseng a=1, the second auxiliary ginseng b=-0.5, the 3rd auxiliary ginseng c=1; If channel matrix conditional number κ > 2.6455, rotation angle θ=0.4636, the first auxiliary ginseng a=1, the second auxiliary ginseng b=-2, the 3rd auxiliary ginseng c=4;

When order of modulation M=16, if channel matrix conditional number κ ∈ [1,2.7575], rotation angle θ=π/4, the first auxiliary ginseng a=1, the second auxiliary ginseng b=-0.5, the 3rd auxiliary ginseng c=1; If channel matrix conditional number κ ∈ (2.7575,6.3293], rotation angle θ=0.4914, the first auxiliary ginseng a=1, the second auxiliary ginseng b=-1.5, the 3rd auxiliary ginseng c=3; If channel matrix conditional number κ ∈ (6.3293,9.7892], rotation angle θ=0.3474, the first auxiliary ginseng a=1, the second auxiliary ginseng b=-2.5, the 3rd auxiliary ginseng c=7; If channel matrix conditional number κ > 9.7892, rotation angle θ=0.2450, the first auxiliary ginseng a=1, the second auxiliary ginseng b=-4, the 3rd auxiliary ginseng c=16;

When order of modulation M=64, if channel matrix conditional number κ≤6.3293, rotation angle θ, the first auxiliary ginseng a, corresponding situation when the value of the second auxiliary ginseng b and the 3rd auxiliary ginseng c is equal to order of modulation M=16; If channel matrix conditional number κ ∈ (6.3293,10.2239], rotation angle θ=0.3474, the first auxiliary ginseng a=1, the second auxiliary ginseng b=-2.5, the 3rd auxiliary ginseng c=7; If channel matrix conditional number κ ∈ (10.2239,13.5809], rotation angle θ=0.5763, the first auxiliary ginseng a=3, the second auxiliary ginseng b=-4.5, the 3rd auxiliary ginseng c=7; If channel matrix conditional number κ ∈ (13.5809,19.2505], rotation angle θ=0.2640, the first auxiliary ginseng a=1, the second auxiliary ginseng b=-3.5, the 3rd auxiliary ginseng c=13; If channel matrix conditional number κ ∈ (19.2505,24.1993], rotation angle θ=0.6235, the first auxiliary ginseng a=7, the second auxiliary ginseng b=-9.5, the 3rd auxiliary ginseng c=13; If channel matrix conditional number κ ∈ (24.1993,27.6122], rotation angle θ=0.3766, the first auxiliary ginseng a=3, the second auxiliary ginseng b=-7.5, the 3rd auxiliary ginseng c=19; If channel matrix conditional number κ ∈ (27.6122,33.3233], rotation angle θ=0.5450, the first auxiliary ginseng a=7, the second auxiliary ginseng b=-11.5, the 3rd auxiliary ginseng c=19; If channel matrix conditional number κ ∈ (33.3233,37.5851], rotation angle θ=0.1757, the first auxiliary ginseng a=1, the second auxiliary ginseng b=-5.5, the 3rd auxiliary ginseng c=31; If channel matrix conditional number κ > 37.5851, rotation angle θ=0.1244, the first auxiliary ginseng a=1, the second auxiliary ginseng b=-8, the 3rd auxiliary ginseng c=64;

When order of modulation M=256, if channel matrix conditional number κ ∈ [1,2.7578], rotation angle θ=π/4, the first auxiliary ginseng a=1, the second auxiliary ginseng b=-0.5, the 3rd auxiliary ginseng c=1; If channel matrix conditional number κ ∈ (2.7578,6.3310], rotation angle θ=0.4914, the first auxiliary ginseng a=1, the second auxiliary ginseng b=-1.5, the 3rd auxiliary ginseng c=3; If channel matrix conditional number κ ∈ (6.3310,10.2277], rotation angle θ=0.3474, the first auxiliary ginseng a=1, the second auxiliary ginseng b=-2.5, the 3rd auxiliary ginseng c=7; If channel matrix conditional number κ ∈ (10.2277,13.5847], rotation angle θ=0.5763, the first auxiliary ginseng a=3, the second auxiliary ginseng b=-4.5, the 3rd auxiliary ginseng c=7; If channel matrix conditional number κ ∈ (13.5847,19.2666], rotation angle θ=0.2640, the first auxiliary ginseng a=1, the second auxiliary ginseng b=-3.5, the 3rd auxiliary ginseng c=13; If channel matrix conditional number κ ∈ (19.2666,24.1797], rotation angle θ=0.6235, the first auxiliary ginseng a=7, the second auxiliary ginseng b=-9.5, the 3rd auxiliary ginseng c=13; If channel matrix conditional number κ ∈ (24.1797,27.5797], rotation angle θ=0.3766, the first auxiliary ginseng a=3, the second auxiliary ginseng b=-7.5, the 3rd auxiliary ginseng c=19; If channel matrix conditional number κ ∈ (27.5797,33.2760], rotation angle θ=0.5450, the first auxiliary ginseng a=7, the second auxiliary ginseng b=-11.5, the 3rd auxiliary ginseng c=19; If channel matrix conditional number κ ∈ (33.2760,38.0613], rotation angle θ=0.1757, the first auxiliary ginseng a=1, the second auxiliary ginseng b=-5.5, the 3rd auxiliary ginseng c=31; If channel matrix conditional number κ ∈ (38.0631,42.5594], rotation angle θ=0.6665, the first auxiliary ginseng a=13, the second auxiliary ginseng b=-16.5, the 3rd auxiliary ginseng c=21; If channel matrix conditional number κ ∈ (42.5594,48.4217], rotation angle θ=0.2766, the first auxiliary ginseng a=3, the second auxiliary ginseng b=-10.5, the 3rd auxiliary ginseng c=37; If channel matrix conditional number κ ∈ (48.4217,51.9294], rotation angle θ=0.1501, the first auxiliary ginseng a=1, the second auxiliary ginseng b=-6.5, the 3rd auxiliary ginseng c=43; If channel matrix conditional number κ ∈ (51.9294,56.4563], rotation angle θ=0.4004, the first auxiliary ginseng a=7, the second auxiliary ginseng b=-16.5, the 3rd auxiliary ginseng c=39; If channel matrix conditional number κ ∈ (56.4563,62.2950], rotation angle θ=0.5234, the first auxiliary ginseng a=13, the second auxiliary ginseng b=-22.5, the 3rd auxiliary ginseng c=39; If channel matrix conditional number κ ∈ (62.2950,65.7926], rotation angle θ=0.3611, the first auxiliary ginseng a=7, the second auxiliary ginseng b=-18.5, the 3rd auxiliary ginseng c=49; If channel matrix conditional number κ ∈ (65.7926,70.3372], rotation angle θ=0.1309, the first auxiliary ginseng a=1, the second auxiliary ginseng b=-7.5, the 3rd auxiliary ginseng c=57; If channel matrix conditional number κ ∈ (70.3372,76.1621], rotation angle θ=0.2178, the first auxiliary ginseng a=3, the second auxiliary ginseng b=-13.5, the 3rd auxiliary ginseng c=61; If channel matrix conditional number κ ∈ (76.1621,81.8983], rotation angle θ=0.5569, the first auxiliary ginseng a=19, the second auxiliary ginseng b=-30.5, the 3rd auxiliary ginseng c=49; If channel matrix conditional number κ ∈ (81.8983,88.8331], rotation angle θ=0.1159, the first auxiliary ginseng a=1, the second auxiliary ginseng b=-8.5, the 3rd auxiliary ginseng c=73; If channel matrix conditional number κ ∈ (88.8331,95.7671], rotation angle θ=0.4148, the first auxiliary ginseng a=13, the second auxiliary ginseng b=-29.5, the 3rd auxiliary ginseng c=67; If channel matrix conditional number κ ∈ (95.7671,100.4424], rotation angle θ=0.2892, the first auxiliary ginseng a=7, the second auxiliary ginseng b=-23.5, the 3rd auxiliary ginseng c=79; If channel matrix conditional number κ ∈ (100.4424,103.8877], rotation angle θ=0.5103, the first auxiliary ginseng a=21, the second auxiliary ginseng b=-37.5, the 3rd auxiliary ginseng c=67; If channel matrix conditional number κ ∈ (103.8877,107.3716], rotation angle θ=0.1040, the first auxiliary ginseng a=1, the second auxiliary ginseng b=-9.5, the 3rd auxiliary ginseng c=91; If channel matrix conditional number κ ∈ (107.3716,110.8182], rotation angle θ=0.1794, the first auxiliary ginseng a=3, the second auxiliary ginseng b=-16.5, the 3rd auxiliary ginseng c=91; If channel matrix conditional number κ ∈ (110.8182,114.3008], rotation angle θ=0.6617, the first auxiliary ginseng a=37, the second auxiliary ginseng b=-47.5, the 3rd auxiliary ginseng c=61; If channel matrix conditional number κ ∈ (114.3008,121.0977], rotation angle θ=0.2676, the first auxiliary ginseng a=7, the second auxiliary ginseng b=-25.5, the 3rd auxiliary ginseng c=93; If channel matrix conditional number κ ∈ (121.0977,128.1588], rotation angle θ=0.3509, the first auxiliary ginseng a=13, the second auxiliary ginseng b=-35.5, the 3rd auxiliary ginseng c=97; If channel matrix conditional number κ ∈ (128.1588,132.7378], rotation angle θ=0.0943, the first auxiliary ginseng a=1, the second auxiliary ginseng b=-10.5, the 3rd auxiliary ginseng c=111; If channel matrix conditional number κ ∈ (132.7378,139.6686], rotation angle θ=0.6125, the first auxiliary ginseng a=39, the second auxiliary ginseng b=-55.5, the 3rd auxiliary ginseng c=79; If channel matrix conditional number κ ∈ (139.6686,145.4671], rotation angle θ=0.4241, the first auxiliary ginseng a=21, the second auxiliary ginseng b=-46.5, the 3rd auxiliary ginseng c=103; If channel matrix conditional number κ ∈ (145.4671,148.3779], rotation angle θ=0.3955, the first auxiliary ginseng a=19, the second auxiliary ginseng b=-45.5, the 3rd auxiliary ginseng c=109; If channel matrix conditional number κ > 148.3779, rotation angle θ=0.0624, the first auxiliary ginseng a=1, the second auxiliary ginseng b=-16, the 3rd auxiliary ginseng c=256;

The 3rd step, calculates orthogonal matrix step: according to the right orthogonal matrix definition of pre-coding matrix F $B=\left[\begin{array}{cc}\cos \mathrm{\&theta;}& -\sin \mathrm{\&theta;}\\ \sin \mathrm{\&theta;}& \cos \mathrm{\&theta;}\end{array}\right]$ Calculate the right orthogonal matrix B of pre-coding matrix F;

The 4th step, calculates diagonal matrix step: according to merit subangle parameter-definition formula calculate merit and divide angular dimensions and according to the diagonal matrix definition of pre-coding matrix F calculate the diagonal matrix Σ of pre-coding matrix F;

The 5th step, calculates pre-coding matrix step: according to pre-coding matrix definition F=V Σ B, calculate pre-coding matrix F, and it is exported as best pre-coding matrix.

The inventive method is owing to utilizing channel matrix conditional number and quadrature amplitude modulation signal order of modulation as indexing parameter, by searching the precoding parameter look-up table that is applicable to 4 rank, rank to 256 quadrature amplitude modulation signals of having constructed in advance, obtain design parameter, according to the closed computing formula design minimum range pre-coding matrix of model parameter, make full use of the limited character property of discrete input signal, utilize grid space feature and correlation theory, on limited lattice point, find the solution of former optimization problem.Compare with existing minimum range method for precoding, the defect of taking this solution route of the inventive method can fundamentally avoid the optimisation strategy based on numerical value exhaustive search to exist, reduce space search scope and search complexity, the parameter look-up table that only need construct by means of calculated off-line can be realized the online real-time design of pre-coding matrix, therefore the inventive method can realize the minimum range method for precoding that is applicable to 4 rank, rank to 256 QAM signals and function admirable with extremely low online computation complexity real-time design, Project Realization flow process is simple, be applicable to such as 802.11n, TD-HSPA+, the new generation broadband wireless such as TD-LTE and TD-LTE-Advanced and mobile communication system.

Accompanying drawing explanation

Fig. 1 is the flow process theory diagram that the present invention is applicable to the minimum range method for precoding of double-flow data transmission.

Fig. 2 is MIMO receiving-transmitting chain signal processing schematic diagram.

Embodiment

Embodiment 1:

The present embodiment be take transmitting terminal, and to adopt 16-QAM modulation system be example, illustrates the operating process that adopts the present invention to be applicable to the minimum range method for precoding of double-flow data transmission.

In the present embodiment, establish MIMO communication system and adopt 16-QAM modulation system to send two paths of data stream, i.e. order of modulation M=16, and total transmitted power P at transmitting terminal ₀=1, interchannel noise variance is made as 1; If two transmitting antennas of transmitting terminal configuration, two reception antennas of receiving terminal configuration, and receiving terminal can directly utilize the channel reciprocity in time division duplex (TDD) system to obtain channel condition information by the complete feeding back channel state information of feedback channel or transmitting terminal, this that makes equivalent virtual channel matrix H is embodied as

$H=\left[\begin{array}{cc}-1.1465+0.3273i& 1.1892-0.1867i\\ 1.1909+0.1746i& -0.0376+0.7258i\end{array}\right].$

Fig. 1 has provided the flow process theory diagram of the minimum range method for precoding of the applicable double-flow data transmission of the present invention.

According to MIMO communication system configuration parameter, transmitting terminal obtain about equivalent virtual channel matrix, order of modulation and power information as input message after, proceed as follows:

The first step, calculates channel matrix conditional number steps A 1: utilize singular value decomposition definition H=U Λ V ^*equivalent virtual channel matrix H is carried out to singular value decomposition, () ^*represent conjugate transpose operation; Obtain the right singular matrix of equivalent virtual channel matrix H $V=\left[\begin{array}{cc}0.8010& -0.5987\\ -0.5315-0.2755i& -0.7112-0.3686i\end{array}\right],$ The diagonal matrix of equivalence virtual channel matrix H $\mathrm{\&Lambda;}=\left[\begin{array}{cc}2.0001& 0\\ 0& 0.9204\end{array}\right],$ Be maximum singular value λ ₁=2.0001, inferior large singular value λ ₂=0.9204; According to matrix conditional number definition κ=λ ₁/ λ ₂can be calculated channel matrix conditional number κ=2.1732;

Second step, parameter finding step A2: according to the Precoding Design parameter look-up table table 1 of realizing an applicable QAM signal of parametric configuration with different channels given below,

Table 1 is applicable to the Precoding Design parameter look-up table of QAM signal

Wherein rotation angle θ can be by anglec of rotation parameter-definition formula calculate and obtain; Utilize the Precoding Design parameter look-up table table 1 of this applicable QAM signal, the interval falling into according to order of modulation M and channel matrix conditional number κ, determines rotation angle θ=π 4 and the first auxiliary ginseng a=1, the second auxiliary ginseng b=-0.5 and the 3rd auxiliary ginseng c=1;

The 3rd step, calculate orthogonal matrix steps A 3: and according to the right orthogonal matrix definition of pre-coding matrix F $B=\left[\begin{array}{cc}\cos \mathrm{\&theta;}& -\sin \mathrm{\&theta;}\\ \sin \mathrm{\&theta;}& \cos \mathrm{\&theta;}\end{array}\right]$ Can calculate the right orthogonal matrix of pre-coding matrix F $B=\frac{\sqrt{2}}{2}\left[\begin{array}{cc}1& -1\\ 1& 1\end{array}\right];$

The 4th step, calculates diagonal matrix steps A 4: according to merit subangle parameter-definition formula can calculate merit minute angular dimensions according to the diagonal matrix definition of pre-coding matrix F can calculate the diagonal matrix of pre-coding matrix F $\mathrm{\&Sigma;}=\left[\begin{array}{cc}0.6233& 0\\ 0& 0.7820\end{array}\right];$

The 5th step, calculates pre-coding matrix steps A 5: according to pre-coding matrix definition F=V Σ B, can calculate pre-coding matrix $F=\left[\begin{array}{cc}0.0220& -0.6841\\ -0.6275-0.3252i& -0.1590-0.0824i\end{array}\right],$ And it is exported as best pre-coding matrix, for transmitting terminal double fluid data-signal is carried out to precoding processing.

Fig. 2 has provided MIMO receiving-transmitting chain signal processing schematic diagram, for characterizing the using method of the simplification system model of the multi-aerial radio communication system that adopts precoder and the best pre-coding matrix of the minimum range method for precoding output that the present invention is applicable to double-flow data transmission.In the information source generation step B1 of transmitting terminal, the data symbol in double-current data s is derived from QAM modulation constellation, and meets unit power: i.e. E[|s| ²]=1, wherein symbol E represents expectation operator, 1 characterizes one 2 * 2 unit matrix; During through precoding step B2, the best pre-coding matrix F that is applicable to the minimum range method for precoding output of double-flow data transmission according to the present invention, carry out the shaping of transmitted signal planisphere, this shaped signal Fs feed-in physical channel, through transmission step B3, make the transmission of transmitted signal channel matrix H, pass through again noise jamming step B4, the multiple Gaussian noise z of stack Cyclic Symmetry, arrive the signal y=HFs+z of receiving terminal, receiving terminal utilizes the final restoring signal that obtains original double-current data of maximum likelihood receiver in maximum likelihood receiver step B5.

Receive the minimum Eustachian distance between two signaling points in signal space can reflect the precoding performance that the minimum range method for precoding that utilizes the present invention to be applicable to double-flow data transmission can reach.Utilize the closed expression formula of minimum Eustachian distance the minimum Eustachian distance that the best pre-coding matrix of computing application the present embodiment output obtains companion matrix wherein $A=\left[\begin{array}{cc}a& b\\ b& c\end{array}\right]=\left[\begin{array}{cc}1& -0.5\\ -0.5& 1\end{array}\right],$ The maximum singular value of companion matrix A is λ _{a, 1}=1.5, time large singular value of companion matrix A is λ _{a, 2}=0.5.Can be calculated minimum Eustachian distance this and the international IEEE-signal of < < are processed transactions > > (IEEE Transactions on Signal Processing, vol.59, no.11, pp.5485 – 5498, Nov.2011) disclosed " a kind of low maximum-likelihood decoding complexity, full diversity gain full rate MIMO method for precoding " (A low ML-decoding complexity, full diversity, full-rate MIMO precoder) the precoding performance that a literary composition obtains is identical.Because pre existing coding method cannot provide the closed expression formula about two angle parameters, can only determine model parameter by numerical search strategy, therefore in identical precoding performance situation, the computation complexity of pre existing coding method is apparently higher than the inventive method, and Project Realization cost is larger.During particularly for high-order QAM modulation signal, numerical search precision also reduces can cause certain performance loss to pre existing coding method.

Although the present embodiment has only adopted the explanation of 16-QAM modulation system to adopt the present invention to be applicable to the operating process of the minimum range method for precoding of double-flow data transmission, but the Precoding Design parameter look-up table that the minimum range method for precoding that is applicable to double-flow data transmission due to the present invention realizes for different channels the applicable QAM signal that parameter constructs has been contained 4 rank, rank to 256 QAM signals, so the present embodiment flow process other any one modulation system of being equally applicable to stipulate in form.The inventive method is equally applicable to number of transmit antennas and reception antenna number and surpasses 2 mimo system configuration mode, need only when the singular value decomposition of utilizing for channel matrix, select to form corresponding to two right singular vectors of channel matrix maximum singular value and time large singular value the right singular matrix V of equivalent virtual channel matrix H.

Compare with the existing relevant design method based on numerical value exhaustive search strategy, the minimum range method for precoding that the present invention is applicable to double-flow data transmission utilizes channel matrix conditional number and quadrature amplitude modulation signal order of modulation as indexing parameter, by searching the precoding parameter look-up table that is applicable to 4 rank, rank to 256 quadrature amplitude modulation signals of having constructed in advance, obtain design parameter, according to the closed computing formula of the anglec of rotation and merit minute angular dimensions, carry out real-time design, it is in fact the limited character property that makes full use of discrete input signal, utilize grid space feature and correlation theory on limited lattice point, to find the solution of former optimization problem, therefore not only have and approach optimum precoding performance, there is extremely low online computation complexity and Project Realization cost simultaneously.Compare with existing minimum range method for precoding, the defect of taking this solution route of the inventive method can fundamentally avoid the optimisation strategy based on numerical value exhaustive search to exist, reduce space search scope and search complexity, the parameter look-up table that only need construct by means of calculated off-line can be realized the online real-time design of pre-coding matrix.Therefore, the present invention can realize the minimum range method for precoding that is applicable to 4 rank, rank to 256 QAM signals and function admirable with extremely low online computation complexity real-time design, Project Realization flow process is simple, is applicable to new generation broadband wireless and mobile communication system such as 802.11n, TD-HSPA+, TD-LTE and TD-LTE-Advanced.

Claims (1)

1. a minimum range method for precoding for applicable double-flow data transmission, establishes transmitting terminal and adopts quadrature amplitude modulation to send two paths of data stream, and total transmitted power is P ₀, the equivalent virtual channel matrix of MIMO flat fading channel and known at transmitting terminal, pre-coding matrix it is characterized in that concrete operation step is:

First utilize singular value decomposition definition H=U Λ V ^*equivalent virtual channel matrix H is carried out to singular value decomposition, and wherein, left singular matrix U and the right singular matrix V of equivalent virtual channel matrix H are unitary matrix, the maximum singular value λ that comprises descending arrangement in the diagonal matrix Λ of equivalent virtual channel matrix H ₁with inferior large singular value λ ₂, V ^*expression is carried out conjugate transpose operation to matrix V; According to matrix conditional number definition κ=λ ₁/ λ ₂calculate channel matrix conditional number κ;

Then rotation angle θ, the first auxiliary ginseng a, the second auxiliary ginseng b and the concrete numerical value of the 3rd auxiliary ginseng c are determined in the interval falling into according to order of modulation M and channel matrix conditional number κ, that is:

When order of modulation M=4, if channel matrix conditional number κ ∈ [1,2.6455], rotation angle θ=π/4, the first auxiliary ginseng a=1, the second auxiliary ginseng b=-0.5, the 3rd auxiliary ginseng c=1; If channel matrix conditional number κ > 2.6455, rotation angle θ=0.4636, the first auxiliary ginseng a=1, the second auxiliary ginseng b=-2, the 3rd auxiliary ginseng c=4;

When order of modulation M=16, if channel matrix conditional number κ ∈ [1,2.7575], rotation angle θ=π/4, the first auxiliary ginseng a=1, the second auxiliary ginseng b=-0.5, the 3rd auxiliary ginseng c=1; If channel matrix conditional number κ ∈ (2.7575,6.3293], rotation angle θ=0.4914, the first auxiliary ginseng a=1, the second auxiliary ginseng b=-1.5, the 3rd auxiliary ginseng c=3; If channel matrix conditional number κ ∈ (6.3293,9.7892], rotation angle θ=0.3474, the first auxiliary ginseng a=1, the second auxiliary ginseng b=-2.5, the 3rd auxiliary ginseng c=7; If channel matrix conditional number κ > 9.7892, rotation angle θ=0.2450, the first auxiliary ginseng a=1, the second auxiliary ginseng b=-4, the 3rd auxiliary ginseng c=16;

When order of modulation M=64, if channel matrix conditional number κ≤6.3293, rotation angle θ, the first auxiliary ginseng a, corresponding situation when the value of the second auxiliary ginseng b and the 3rd auxiliary ginseng c is equal to order of modulation M=16; If channel matrix conditional number κ ∈ (6.3293,10.2239], rotation angle θ=0.3474, the first auxiliary ginseng a=1, the second auxiliary ginseng b=-2.5, the 3rd auxiliary ginseng c=7; If channel matrix conditional number κ ∈ (10.2239,13.5809], rotation angle θ=0.5763, the first auxiliary ginseng a=3, the second auxiliary ginseng b=-4.5, the 3rd auxiliary ginseng c=7; If channel matrix conditional number κ ∈ (13.5809,19.2505], rotation angle θ=0.2640, the first auxiliary ginseng a=1, the second auxiliary ginseng b=-3.5, the 3rd auxiliary ginseng c=13; If channel matrix conditional number κ ∈ (19.2505,24.1993], rotation angle θ=0.6235, the first auxiliary ginseng a=7, the second auxiliary ginseng b=-9.5, the 3rd auxiliary ginseng c=13; If channel matrix conditional number κ ∈ (24.1993,27.6122], rotation angle θ=0.3766, the first auxiliary ginseng a=3, the second auxiliary ginseng b=-7.5, the 3rd auxiliary ginseng c=19; If channel matrix conditional number κ ∈ (27.6122,33.3233], rotation angle θ=0.5450, the first auxiliary ginseng a=7, the second auxiliary ginseng b=-11.5, the 3rd auxiliary ginseng c=19; If channel matrix conditional number κ ∈ (33.3233,37.5851], rotation angle θ=0.1757, the first auxiliary ginseng a=1, the second auxiliary ginseng b=-5.5, the 3rd auxiliary ginseng c=31; If channel matrix conditional number κ > 37.5851, rotation angle θ=0.1244, the first auxiliary ginseng a=1, the second auxiliary ginseng b=-8, the 3rd auxiliary ginseng c=64;

When order of modulation M=256, if channel matrix conditional number κ ∈ [1,2.7578], rotation angle θ=π/4, the first auxiliary ginseng a=1, the second auxiliary ginseng b=-0.5, the 3rd auxiliary ginseng c=1; If channel matrix conditional number κ ∈ (2.7578,6.3310], rotation angle θ=0.4914, the first auxiliary ginseng a=1, the second auxiliary ginseng b=-1.5, the 3rd auxiliary ginseng c=3; If channel matrix conditional number κ ∈ (6.3310,10.2277], rotation angle θ=0.3474, the first auxiliary ginseng a=1, the second auxiliary ginseng b=-2.5, the 3rd auxiliary ginseng c=7; If channel matrix conditional number κ ∈ (10.2277,13.5847], rotation angle θ=0.5763, the first auxiliary ginseng a=3, the second auxiliary ginseng b=-4.5, the 3rd auxiliary ginseng c=7; If channel matrix conditional number κ ∈ (13.5847,19.2666], rotation angle θ=0.2640, the first auxiliary ginseng a=1, the second auxiliary ginseng b=-3.5, the 3rd auxiliary ginseng c=13; If channel matrix conditional number κ ∈ (19.2666,24.1797], rotation angle θ=0.6235, the first auxiliary ginseng a=7, the second auxiliary ginseng b=-9.5, the 3rd auxiliary ginseng c=13; If channel matrix conditional number κ ∈ (24.1797,27.5797], rotation angle θ=0.3766, the first auxiliary ginseng a=3, the second auxiliary ginseng b=-7.5, the 3rd auxiliary ginseng c=19; If channel matrix conditional number κ ∈ (27.5797,33.2760], rotation angle θ=0.5450, the first auxiliary ginseng a=7, the second auxiliary ginseng b=-11.5, the 3rd auxiliary ginseng c=19; If channel matrix conditional number κ ∈ (33.2760,38.0613], rotation angle θ=0.1757, the first auxiliary ginseng a=1, the second auxiliary ginseng b=-5.5, the 3rd auxiliary ginseng c=31; If channel matrix conditional number κ ∈ (38.0631,42.5594], rotation angle θ=0.6665, the first auxiliary ginseng a=13, the second auxiliary ginseng b=-16.5, the 3rd auxiliary ginseng c=21; If channel matrix conditional number κ ∈ (42.5594,48.4217], rotation angle θ=0.2766, the first auxiliary ginseng a=3, the second auxiliary ginseng b=-10.5, the 3rd auxiliary ginseng c=37; If channel matrix conditional number κ ∈ (48.4217,51.9294], rotation angle θ=0.1501, the first auxiliary ginseng a=1, the second auxiliary ginseng b=-6.5, the 3rd auxiliary ginseng c=43; If channel matrix conditional number κ ∈ (51.9294,56.4563], rotation angle θ=0.4004, the first auxiliary ginseng a=7, the second auxiliary ginseng b=-16.5, the 3rd auxiliary ginseng c=39; If channel matrix conditional number κ ∈ (56.4563,62.2950], rotation angle θ=0.5234, the first auxiliary ginseng a=13, the second auxiliary ginseng b=-22.5, the 3rd auxiliary ginseng c=39; If channel matrix conditional number κ ∈ (62.2950,65.7926], rotation angle θ=0.3611, the first auxiliary ginseng a=7, the second auxiliary ginseng b=-18.5, the 3rd auxiliary ginseng c=49; If channel matrix conditional number κ ∈ (65.7926,70.3372], rotation angle θ=0.1309, the first auxiliary ginseng a=1, the second auxiliary ginseng b=-7.5, the 3rd auxiliary ginseng c=57; If channel matrix conditional number κ ∈ (70.3372,76.1621], rotation angle θ=0.2178, the first auxiliary ginseng a=3, the second auxiliary ginseng b=-13.5, the 3rd auxiliary ginseng c=61; If channel matrix conditional number κ ∈ (76.1621,81.8983], rotation angle θ=0.5569, the first auxiliary ginseng a=19, the second auxiliary ginseng b=-30.5, the 3rd auxiliary ginseng c=49; If channel matrix conditional number κ ∈ (81.8983,88.8331], rotation angle θ=0.1159, the first auxiliary ginseng a=1, the second auxiliary ginseng b=-8.5, the 3rd auxiliary ginseng c=73; If channel matrix conditional number κ ∈ (88.8331,95.7671], rotation angle θ=0.4148, the first auxiliary ginseng a=13, the second auxiliary ginseng b=-29.5, the 3rd auxiliary ginseng c=67; If channel matrix conditional number κ ∈ (95.7671,100.4424], rotation angle θ=0.2892, the first auxiliary ginseng a=7, the second auxiliary ginseng b=-23.5, the 3rd auxiliary ginseng c=79; If channel matrix conditional number κ ∈ (100.4424,103.8877], rotation angle θ=0.5103, the first auxiliary ginseng a=21, the second auxiliary ginseng b=-37.5, the 3rd auxiliary ginseng c=67; If channel matrix conditional number κ ∈ (103.8877,107.3716], rotation angle θ=0.1040, the first auxiliary ginseng a=1, the second auxiliary ginseng b=-9.5, the 3rd auxiliary ginseng c=91; If channel matrix conditional number κ ∈ (107.3716,110.8182], rotation angle θ=0.1794, the first auxiliary ginseng a=3, the second auxiliary ginseng b=-16.5, the 3rd auxiliary ginseng c=91; If channel matrix conditional number κ ∈ (110.8182,114.3008], rotation angle θ=0.6617, the first auxiliary ginseng a=37, the second auxiliary ginseng b=-47.5, the 3rd auxiliary ginseng c=61; If channel matrix conditional number κ ∈ (114.3008,121.0977], rotation angle θ=0.2676, the first auxiliary ginseng a=7, the second auxiliary ginseng b=-25.5, the 3rd auxiliary ginseng c=93; If channel matrix conditional number κ ∈ (121.0977,128.1588], rotation angle θ=0.3509, the first auxiliary ginseng a=13, the second auxiliary ginseng b=-35.5, the 3rd auxiliary ginseng c=97; If channel matrix conditional number κ ∈ (128.1588,132.7378], rotation angle θ=0.0943, the first auxiliary ginseng a=1, the second auxiliary ginseng b=-10.5, the 3rd auxiliary ginseng c=111; If channel matrix conditional number κ ∈ (132.7378,139.6686], rotation angle θ=0.6125, the first auxiliary ginseng a=39, the second auxiliary ginseng b=-55.5, the 3rd auxiliary ginseng c=79; If channel matrix conditional number κ ∈ (139.6686,145.4671], rotation angle θ=0.4241, the first auxiliary ginseng a=21, the second auxiliary ginseng b=-46.5, the 3rd auxiliary ginseng c=103; If channel matrix conditional number κ ∈ (145.4671,148.3779], rotation angle θ=0.3955, the first auxiliary ginseng a=19, the second auxiliary ginseng b=-45.5, the 3rd auxiliary ginseng c=109; If channel matrix conditional number κ > 148.3779, rotation angle θ=0.0624, the first auxiliary ginseng a=1, the second auxiliary ginseng b=-16, the 3rd auxiliary ginseng c=256;

Again according to the right orthogonal matrix definition of pre-coding matrix F $B=\left[\begin{array}{cc}\cos \mathrm{\&theta;}& -\sin \mathrm{\&theta;}\\ \sin \mathrm{\&theta;}& \cos \mathrm{\&theta;}\end{array}\right]$ Calculate the right orthogonal matrix B of pre-coding matrix F;

According to merit subangle parameter-definition formula calculate merit and divide angular dimensions and according to the diagonal matrix definition of pre-coding matrix F calculate the diagonal matrix Σ of pre-coding matrix F;

Finally according to pre-coding matrix definition F=V Σ B, calculate pre-coding matrix F, it is exported as best pre-coding matrix.