CN1835425A - Self-adaptive modulation method based on multi-user precode - Google Patents

Abstract

Purpose of the invention is to raise performance of anti- error code in multi-user MIMO system. The method includes steps: base station obtains channel matrix and noise power of each user's terminal; using channel matrix and noise power of each user's terminal and prearranged precoding method calculates equivalent SNR of each user's terminal; determining self adaptive modulation parameter table from equivalent SNR; using water flooding method determines value of power distribution based on equivalent SNR of each user's terminal; calculating new SNR of each user's terminal from noise power and value of power distribution of each user's terminal, and determining corresponding mode of modulation from the self adaptive modulation parameter table based on new SNR; redistributing power for users' terminals in same modulation mode according to mode in same SNR so as to obtain final distributed power for each user's terminal, sending data from antenna based on final distributed power and mode.

Description

Self-adaptive modulation method based on multi-user pre-coding

Technical field

The present invention relates in the multiuser MIMO wireless communication system to be used for improving the error-resilient performance of mimo system based on the self-adaptive modulation method and the equipment of precoding.

Background technology

Multiple-input, multiple-output (MIMO) technology is the important breakthrough of wireless mobile communications art.The MIMO technology is meant the transmission of data and receives and all adopted many antennas.Studies show that utilize the MIMO technology can improve the capacity of channel, the while also can be improved the reliability of channel, reduces the error rate.The MIMO technology has great potentiality for the capacity that improves wireless communication system, is the key technology that the third generation mobile communication system adopts.Multiuser mimo communication is one of current research focus.

Figure 7 shows that the Single User MIMO system configuration schematic diagram of common employing.In this structure, make a start and receiving end adopt n respectively _TAnd n _RIndividual antenna carries out the transmission and the reception of signal.At transmitting terminal, data to be sent are at first passed through the processing of serial to parallel conversion module 701, are divided into n _TIndividual data flow, each data flow is launched from the transmitting antenna 702 of correspondence.At receiving terminal, at first by n _R Individual reception antenna 703 receives signal, carries out channel estimating by channel estimation module 704 according to this received signal then, estimates current characteristic of channel matrix H.Detector 705 utilizes this characteristic of channel matrix H to detect to received signal, demodulates the information bit of making a start and sending.

What Fig. 7 provided is the structure chart of Single User MIMO, and the structure chart of multiuser MIMO as shown in Figure 1.In Fig. 1, the base station transmits to user terminal 1 and user terminal 2 simultaneously.But the desired signal of user terminal 2 has caused interference to user terminal 1, and the desired signal of user terminal 1 has caused interference to user terminal 2.And user terminal 1 can't obtain the characteristic of channel of user terminal 2, and user terminal 2 can't obtain the characteristic of channel of user terminal 1 equally.

[multi-user MIMO system model]

Multi-user's model as shown in the formula:

x＝Hs+n (1)

In formula (1), n represents that the average on the reception antenna is zero, variance is σ ²White Gaussian noise, x is the signal vector of reception antenna, s is the transmission signal after the precoding, H is a channel matrix.

[traditional multi-user MIMO pre-encoding method]

1, channel reverse method for precoding

Fig. 8 has provided traditional inverse channel precoding multiuser MIMO method.In Fig. 8, data d is earlier through the channel reverse matrix H ^-1Handle, behind channel H, H ^-1H=I has eliminated interference fully, but its cost is in order to eliminate interference, to have improved transmitting power, or under the constant situation of transmitting power, equivalence has improved noise.

2, Tomlinson-Harashima multi-user MIMO pre-encoding method

The Tomlinson-Harashima multi-user pre-coding method as shown in Figure 9.The variant LQ that this method utilizes QR to decompose decomposes, and channel matrix decomposition H is as follows:

H＝SF ^H (2)

In the formula (2), S is a lower triangular matrix, and F is unitary matrice (F ^HF=I).Definition then

C＝GHF＝GS (3)

In the formula (3) $G=\mathrm{diag}(s_{11}^{-1},...,s_{\mathrm{KK}}^{-1}),$ C is that a diagonal is 1 lower triangular matrix.The implementation procedure of this method as shown in Figure 9.The thought of this method is, if known disturbances can deduct in advance at transmit leg.The effect of operation C-I in Fig. 9 just is this.C-I is that a diagonal is 0 lower triangular matrix.Concerning first user, do not disturb, the 2nd user will deduct first user's interference, and K user will deduct 1 interference of arriving K-1 user.

MOD among Fig. 9 represents to ask modular arithmetic, can reduce transmitting power like this.When K user obtains the data of oneself, received signal be multiply by K the diagonal entry g of G _KRecover, then by asking mould and demodulation output.This method for precoding is owing to reduced transmitting power, and performance is better than the method for matrix inversion, still, send gain control value g to each user in each frame data _K, need extra expense.

3, reduce the method for precoding of (lattice reduction) based on grid

Based on the method for precoding of lattice reduction is following H is carried out the pseudo inverse matrix decomposition:

H _red＝H ^-1×T (4)

H _RedRespectively classify the base vector of nearly orthogonal as.T and T ^-1Each element is an integer, and determinant is ± 1.

The characteristic of utilizing T then is in conjunction with asking modular arithmetic to obtain precoding.

s＝H _redT(T ^-1d?modτ) (5)

In the formula (5), d is the symbol before the precoding, and τ is a boundary value, and the real part of constellation point and imaginary part are between-0.5 τ and 0.5 τ.

4, traditional AMC method for precoding

The basic principle of AMC method is to change the form of modulation and coding and it is adapted with channel condition in the system constraint scope, and channel condition then can be estimated by sending feedback.In the AMC system, the general user uses the modulation system of higher-order and higher code rate under desirable channel condition, and at the next modulation coding mode with lower-order of less-than-ideal channel condition.

Summary of the invention

The objective of the invention is to, a kind of self-adaptive modulation method based on multi-user pre-coding is provided, to improve the descending error performance of multiuser MIMO.

In one aspect of the invention, proposed a kind of self-adaptive modulation method based on multi-user pre-coding, may further comprise the steps: the base station obtains the channel matrix and the noise power of each user terminal; Calculate the equivalent signal-to-noise ratio of each user terminal by channel matrix, noise power and the predetermined method for precoding of each user terminal; Equivalent signal-to-noise ratio by each user terminal is determined the adaptive modulation parameter table; According to the equivalent signal-to-noise ratio of each user terminal, determine the power division value of each user terminal with water-filling method; From the noise power and the power division value of each user terminal, calculate the new signal to noise ratio of each user terminal, and from the adaptive modulation parameter table, determine corresponding modulation system for each user terminal according to new signal to noise ratio; In the mode that signal to noise ratio is identical the user terminal that belongs to same modulation system is redistributed power, obtain the final distribution power of each user terminal; Based on the final distribution power and the modulation system of each user terminal, from the data of each user terminal of antenna transmission.

Utilize method of the present invention, can improve error-resilient performance in the multi-user MIMO system.

Description of drawings

Fig. 1 is the schematic diagram according to the multi-user MIMO system of the embodiment of the invention;

Fig. 2 is the structure chart according to the mimo system of the embodiment of the invention;

Fig. 3 is the flow chart according to the self-adaptive modulation method of the embodiment of the invention;

Fig. 4 shows the M-QAM modulation constellation;

The performance that Fig. 5 has provided diverse ways compares (16-QAM, 4 * 4);

Fig. 6 has provided the comparison (12 bits of handling up) that has or not Adaptive Modulation;

Fig. 7 is a Single User MIMO system configuration schematic diagram;

Fig. 8 is traditional inverse channel precoding multiuser MIMO method; And

Fig. 9 be traditional Tomlinson-Harashima multi-user pre-coding method.

Embodiment

The present invention will be described in detail with the form of embodiment for the contrast accompanying drawing below.

Fig. 2 shows the structure chart according to the mimo system of the embodiment of the invention.

The applied mimo system of the present invention generally includes base station 10 and a plurality of user terminal 201-20N, because the formation of each terminal is roughly the same, therefore only describes at user terminal 201 here.

Base station 10 comprises data source 101, AM module 1021-102N, precoding module 1031 and the transmitting antenna 1041-104N that has stored each user's data such as memory.User terminal 201 comprises reception antenna 2011, channel and Noise Estimation module 2012 and demodulation module 2013.Equally, user terminal 20N comprises reception antenna 201N, channel and Noise Estimation module 20N2 and demodulation module 20N3.

At first, base station 10 obtains channel matrix H and noise power according to each user's feedback channel 301-30N.Above-mentioned noise power is to be estimated by the channel of user terminal and noise power estimation module 2012, and by feedback channel 301 noise power of this user terminal has been issued Adaptive Modulation (AM) module 1021 of base station 10, channel matrix H has been sent to the precoding module 1031 of base station 10.Then, the form of Adaptive Modulation of each user terminals and the transmitted power of each user terminal are determined in base station 10 by Adaptive Modulation module 1031 according to channel matrix H, noise power.

After the modulation system of having determined each user terminal and transmitted power, self adaptation adjusting module 1021 is issued precoding module 1021 from the data to be sent of data source 101 each user terminals of acquisition with these data.According to corresponding modulation system the data of each user terminal are launched from transmitting antenna 1031 then.

After each user terminal is received data by reception antenna 2011-20N1, on the one hand,, give base station 10 with these feedback information by feedback channel 301-30N by the noise power that channel and noise power estimation module 2012-20N2 estimate the signal of reception.Action above repeating.On the other hand, the signal that receives is offered demodulation module 2013-20N3, demodulation module carries out demodulation to the data that receive then, and the output demodulated data is to other module (not shown).

Fig. 3 shows the flow chart according to the self-adaptive modulation method of the embodiment of the invention.

After flow process began, at step S301, base station 10 obtained the channel matrix and the noise power of user terminal.For time division duplex (TDD) system, because the symmetry of up-downgoing, channel matrix can be by being directly used in the precoding of down channel to the resulting channel of up channel estimating.For Frequency Division Duplexing (FDD) (FDD) system, need the user to give the base station with the feedback information of channel.In addition, the channel of user terminal 201 and Noise Estimation module are estimated the noise power of the signal of reception, and send it to the AM module of base station 10 by feedback channel 301.Like this, base station 10 has obtained the channel matrix H and the noise power of each user terminal.

At step S302, base station 10 calculates each user's equivalent signal-to-noise ratio according to channel matrix H, as in the method for the described precoding of background technology part and the noise power of each user terminal.For example, be the method for channel reverse such as the method for precoding that adopts, suppose that the transmission symbol before the precoding is d, the transmission symbol after the precoding is s, then

s＝H ^-1d (6)

Suppose the transmission symbol d normalization that precoding is preceding, be shown below

E(dd ^H)＝I (7)

Then according to (7),

E(ss ^H)＝H ^-1H ^-H (8)

Formula (7) shows that the average transmit power of i root antenna becomes H by 1 ^-1Capable 2 norms of i square, be equivalent to improve transmitting power.The noise power of supposing i user is σ _i ², H ^-1Capable 2 norms of i square be β _i, then can calculate i user's equivalent signal-to-noise ratio according to following formula:

${\mathrm{SNR}}_i=\frac{1}{{\mathrm{\β}}_i{\mathrm{\σ}}_i^2}---\left(9\right)$

After the equivalent signal-to-noise ratio of having determined each user, flow process enters step S303, determines possibility modulation system accordingly according to each user's equivalent signal-to-noise ratio, forms the parameter list of an Adaptive Modulation.Usually, can select different modulation systems at different signal to noise ratios, for example, be lower than in signal to noise ratio under the situation of 5dB, can select the QPSK modulation system, signal to noise ratio greater than the situation of 5dB less than 10dB under, select 8PSK modulation mode, under the situation of signal to noise ratio greater than 10dB, select the 16QAM modulation system, or the like.After having determined modulation system, relevant parameters table, for example transmitting power or the like are all arranged at a kind of modulation system.

Next, at step S304, obtain each user's equivalent signal-to-noise ratio in base station 10 after, remove characteristics of interference according to precoding, with the method for water filling, determine the power division value of each user terminal.

The method of water filling is meant, for the individual independently parallel channel of n, reaches maximum channel capacity by each user being distributed the method for power.For example, total power limited in

$\underset{i=1}{\overset{n}{\mathrm{\Σ}}}{P}_i=P---\left(10\right)$

Situation under, the gain of establishing each channel is λ _i, the noise power of each channel is σ _i ², then the signal to noise ratio of each channel is

${\mathrm{SNR}}_i=\frac{P_i{\mathrm{\λ}}_i}{{\mathrm{\σ}}_i^2}---\left(11\right)$

Its channel capacity is

$C=\underset{i=1}{\overset{n}{\mathrm{\Σ}}}{\log }_2(1+{\mathrm{SNR}}_i)=\underset{i=1}{\overset{n}{\mathrm{\Σ}}}{\log }_2(1+\frac{P_i{\mathrm{\λ}}_i}{{\mathrm{\σ}}_i^2})---\left(12\right)$

According to the method for Lagrange multiplier, introduce auxiliary function

$Z=\underset{i=1}{\overset{n}{\mathrm{\Σ}}}{\log }_2(1+\frac{P_i{\mathrm{\λ}}_i}{{\mathrm{\σ}}_i^2})+L(P-\underset{i=1}{\overset{n}{\mathrm{\Σ}}}{P}_i)---\left(13\right)$

In formula (13), L is the Lagrange multiplier.Make partial derivative

$\frac{\∂Z}{{\∂P}_i}=\frac{1}{\ln 2}\frac{{\mathrm{\λ}}_i/{\mathrm{\σ}}_i^2}{1+P_i{\mathrm{\λ}}_i/{\mathrm{\σ}}_i^2}-L=0---\left(14\right)$

Can get

$P_i=u-\frac{{\mathrm{\σ}}_i^2}{{\mathrm{\λ}}_i}---\left(15\right)$

Early in the formula (15), u is a constant, and its value is 1/ (Lln2).And P _iSatisfy power constraint (10), and P _iMore than or equal to 0.

In fact, because precoding by linear transformation, when the reception of each user terminal, has been eliminated interference fully.Like this, the recipient, be exactly parallel no interchannel interference channel.According to above-mentioned water-filling method, the channel that signal to noise ratio is high should distribute higher power, to obtain more high power capacity.Therefore, it is feasible distributing power with the method for water filling according to the equivalent signal-to-noise ratio of each user terminal.Total transmitting power is N * P (average transmit power that is each transmitting antenna is P), and then the power division to i user is

$P\left(i\right)={(\frac{1}{\mathrm{\λ}}-\frac{1}{{\mathrm{SNR}}_i})}^{+}---\left(16\right)$

In the formula (16), SNR _iBe i user's equivalent signal-to-noise ratio, x ⁺Be defined as x ⁺=max (x, 0), λ satisfies formula (17).

$\frac{1}{N}\underset{m=1}{\overset{N}{\mathrm{\Σ}}}{(\frac{1}{\mathrm{\λ}}-\frac{1}{{\mathrm{SNR}}_m})}^{+}=P---\left(17\right)$

At step S305, because the noise power of each user terminal is known, can calculate the new signal to noise ratio snr of each user terminal this moment according to from above-mentioned power division value '.Then, according to this SNR ', from the adaptive modulation parameter table, select to obtain desired modulation system.Then, to a kind of modulation system again according to or etc. the principle of signal to noise ratio redistribute power.

For example, the number of transmit antennas of supposing base station 10 is N, and each user terminal has an antenna, and the number of user terminal also is N, and then H is the matrix of N * N.

After modulation system has been determined, will redistribute power with a kind of modulation system as one group.Such as, it all is 16-QAM that k user arranged, giving the power of their distribution is respectively P (1), and P (2) ..., P (k).These 16 users' approximate signal to noise ratio all satisfies the lower limit that uses 16-QAM.In the process of power division, the power that the user that signal to noise ratio is high more distributes is with high, and its accessible capacity is also high more.But correspond on the Adaptive Modulation, have only optional several modulation system.And in the set with a kind of modulation system, P (1), P (2) ..., the power that P (k) distributes is also different, and just the power of the distribution that signal to noise ratio is high is higher.Like this, just cause certain user's performance fine (power of distribution is very high), and certain user's poor-performing (power of distribution is relatively low), cause the imbalance of performance.So, in the set of same modulation, redistribute the gross power in this set.Signal to noise ratio principles such as its rule is, as described below.

Etc. the signal to noise ratio principle:

Order at k user's of this modulation system total allocation power is:

$P_{M-\mathrm{QAM}}=\underset{i=1}{\overset{k}{\mathrm{\Σ}}}P\left(i\right)---\left(18\right)$

Principle etc. signal to noise ratio is that the P (i) ' that redistributes is satisfied

$\frac{P{\left(1\right)}^{\′}}{{\mathrm{\σ}}_1^2}=\frac{P{\left(2\right)}^{\′}}{{\mathrm{\σ}}_2^2}=\·\·\·=\frac{P{\left(k\right)}^{\′}}{{\mathrm{\σ}}_k^2}---\left(19\right)$

(19) in, satisfy $P_{M-\mathrm{QAM}}=\underset{i=1}{\overset{k}{\mathrm{\Σ}}}P{\left(i\right)}^{\′}.$

This can be regarded as Adaptive Modulation and combining that power is controlled.Adaptive Modulation is to make the high user of signal to noise ratio distribute more bits according to the principle of water filling.But after modulation scheme was determined, the usefulness of same modulation system was indoor, and power will divide according to fairness, like this, in same modulation system, each user's received signal to noise ratio basically identical, can improve total error performance, avoid artificial causing " near-far interference ".

At step S306, after modulation system that obtains each user terminal and power division value, carry out precoding according to method for multi-user pre-coding.The method for precoding here can adopt the described method for precoding of background technology part in the application.

At last, at step S307, the data that obtain are sent according to power division value and corresponding modulation system from antenna 1041-104N.

In emulation, the number of antennas of base station 10 is 4, and the number of user terminal also is 4, antenna of each user.Fig. 5 has provided method, the method for Tomlinson-Harashima precoding and the comparison of lattice reduction method with channel reverse (ZF).As can be seen from the figure, the method for precoding of Lattice reduction is wanted excellent preceding two kinds.

Fig. 6 has provided (on average each user's 3 bit) when handling up 12 bits, the comparison when having or not Adaptive Modulation, and SNR represents among Fig. 6 is the ratio of the average noise power of the average transmitting power of precoding and receiving terminal.Among Fig. 6, expression be no Adaptive Modulation the time, 8PSK, the performance of ZF, the comparison of the performance of ZF-AM and the performance of LR-AM.Wherein, (mode of modulation is at 16QAM for the performance that has been to use the Adaptive Modulation among the present invention that ZF-AM represents, select among the QPSK, 4 total bit numbers of user are 12), LR-AM is to use the performance of Adaptive Modulation of the present invention and Lattice reduction.Therefrom as can be seen, used Adaptive Modulation of the present invention, performance improves a lot.

The above; only be a kind of embodiment among the present invention, but protection scope of the present invention is not limited thereto, anyly is familiar with the people of this technology in the disclosed technical scope of the present invention; the conversion that can expect easily or replacement all should be encompassed in of the present invention comprising within the scope.Therefore, protection scope of the present invention should be as the criterion with the protection range of claims.

Claims (7)

1, a kind of self-adaptive modulation method based on multi-user pre-coding is characterized in that, may further comprise the steps:

The base station obtains the channel matrix and the noise power of each user terminal;

Calculate the equivalent signal-to-noise ratio of each user terminal by channel matrix, noise power and the predetermined method for precoding of each user terminal;

Equivalent signal-to-noise ratio by each user terminal is determined the adaptive modulation parameter table;

According to the equivalent signal-to-noise ratio of each user terminal, determine the power division value of each user terminal with water-filling method;

From the noise power and the power division value of each user terminal, calculate the new signal to noise ratio of each user terminal, and from the adaptive modulation parameter table, determine corresponding modulation system for each user terminal according to new signal to noise ratio;

In the mode that signal to noise ratio is identical the user terminal that belongs to same modulation system is redistributed power, obtain the final distribution power of each user terminal;

Based on the final distribution power and the modulation system of each user terminal, from the data of each user terminal of antenna transmission.

2, the method for claim 1 is characterized in that, under the situation of time division duplex, the base station utilizes the symmetry of channel, obtains channel matrix.

3, the method for claim 1 is characterized in that, under the situation of Frequency Division Duplexing (FDD), obtains channel matrix from the feedback channel from each user terminal.

4, the method for claim 1 is characterized in that, described predetermined method for precoding is the method for channel reverse, the equivalent signal-to-noise ratio of following each user terminal of calculating:

${\mathrm{SNR}}_i=\frac{1}{{\mathrm{\β}}_i{\mathrm{\σ}}_i^2}$

σ wherein _i ²Be the noise power of i user terminal, β _iBe the inverse matrix H of channel matrix ^-1Capable 2 norms of i square.

5, the method for claim 1 is characterized in that, described water-filling method is meant, for the individual independently parallel channel of n, by distributing power to reach maximum channel capacity to each user terminal.

6, the method for claim 1 is characterized in that, following the user terminal that belongs to same modulation system is redistributed power:

Total allocation power the k that belongs to a same modulation system user is $P_{M-\mathrm{QAM}}=\underset{i=1}{\overset{k}{\mathrm{\Σ}}}P\left(i\right)$ Situation under, make the power P (i) redistributed ' satisfy

$\frac{P{\left(1\right)}^{\′}}{{\mathrm{\σ}}_1^2}=\frac{{P\left(2\right)}^{\′}}{{\mathrm{\σ}}_2^2}=...=\frac{{P\left(k\right)}^{\′}}{{\mathrm{\σ}}_k^2}$

And $P_{M-\mathrm{QAM}}=\underset{i=1}{\overset{k}{\mathrm{\Σ}}}{P\left(i\right)}^{\′}.$

7, as the described method of one of claim 1-6, it is characterized in that, described modulation system comprise 16QAM, 8PSK and QPSK one of at least.

Images