CN102170412A - Interruption relaxation alignment method based on multi-point multi-user coordination downlink - Google Patents

Abstract

The invention discloses an interruption relaxation alignment method based on multi-point multi-user coordination downlink. The method specially comprises the following steps: (1) completing channel estimation at a sending end by using a corresponding channel estimation technique according to that different performances and transmission modes are required by communication systems to obtain channel matrix parameters; (2) establishing an optimized equation consisting of constraint conditions and a target function, wherein the constraint conditions are formed after relaxed interruption and alignment are carried out on constraint conditions by a relaxation factor lambda; (3) solving the optimized equation; (4) respectively performing premultiplication on a sending signal by various receiving ends with a solved wave beam formation matrix so as to complete pretreatment; and (5) carrying out conjugated transposition premultiplication of the received signal by various receiving ends with a solved interruption elimination matrix so as to complete post-treatment. The method has the advantages that a wave beam formation matrix set and an interruption elimination matrix set which are used for inhibiting multi-user interruption can be rapidly solved, and the pre-treatment and post-treatment are respectively carried out by the sending end and receiving ends so as to inhibit the multi-user interruption.

Description

Based on the multiple spot multi-user lax alignment schemes of descending interference of cooperating

Technical field

The invention belongs to communication technical field, further relating to the multiple spot multi-user cooperates in the downlink communication technical field, based on disturbing on the alignment techniques basis, adopt and disturb relaxing techniques, the transmitting terminal signal is carried out preliminary treatment and the receiving terminal received signal is carried out reprocessing, so that suppress the multi-user interference that produces when the multiple spot multi-user cooperates downlink communication.

Background technology

At the down link of multiple spot multi-user cooperation communication system, a plurality of transmission points send information to a plurality of users simultaneously by cooperation way.Any one user also will receive from the signal of other transmission point when receiving useful signal, may produce multi-user interference.Disturbing for eliminating, mainly is to adopt to disturb alignment techniques to make expectation received signal and interference signal obviously can divide at present, reaches the purpose of eliminating multi-user interference.

Krishna Gomadam is at article " Approaching the Capacity of Wireless Networksthrough Distributed Interference Alignment " (Global TelecommunicationsConference, 2008.IEEE GLOBECOM 2008.IEEE, pp.1-6, Nov.30 2008-Dec.42008) united beam was shaped and realizes disturbing the method for aliging when middle a kind of multiple spot multi-user of proposition cooperated downlink transfer.The implementation step of this method is: first, the channel information of all links is taken all factors into consideration at the Combined Treatment center of transmitting terminal, based on the optimization aim that makes channel capacity maximization (promptly the equivalent independent channel that obtains of all users is counted the sum maximum), for each transmitting terminal and receiving terminal calculate optimum beam shaping matrix and interference cancellation matrix; The second, transmitting terminal carries out preliminary treatment (being that beam shaping matrix premultiplication sends signal) to signal before sending signal; The 3rd, receiving terminal carries out reprocessing (being the signal that the conjugate transpose premultiplication of interference cancellation matrix receives) to received signal.Though this method has reached the purpose of eliminating multi-user interference by the Combined Treatment of transmitting terminal and receiving terminal, but, still there is weak point in this method: when transmitting terminal number and receiving terminal number all more than or equal to 3 the time, to finding the solution of beam shaping of associating is a NP-difficulty (Nondeterministic Polynomial-time hard, the nondeterministic polynomial time trouble) problem, be difficult in the polynomial time system of solving and be used to eliminate required beam shaping set of matrices of multi-user interference and interference cancellation matrix set, the united beam that causes transmitting terminal and receiving terminal on the engineering at all to have no idea to carry out next step is shaped, and can't realize disturbing alignment to eliminate multi-user interference.Although can adopt the iterative method to obtain beam shaping set of matrices and interference cancellation matrix set, in calculating process, multi-user interference is exaggerated, poor effect in the practical operation simultaneously.

Summary of the invention

The objective of the invention is to overcome the deficiency of above-mentioned prior art, a kind of united beam manufacturing process based on " disturbing lax alignment " is proposed, that can seek apace that united beam is shaped effectively separates, and realizes disturbing lax alignment to suppress the multiple spot multi-user multi-user interference that downlink communication produces of cooperating.

The concrete steps that the present invention realizes comprise as follows:

(1) obtains the channel matrix parameter.According to communication system different to the difference of performance requirement and transmission means, transmitting terminal adopts corresponding channel estimation technique to finish channel estimating, obtains the channel matrix parameter.

(2) set up the optimization equation.

2a) determine constraints.According to the remaining multi-user interference of each system, determine corresponding relaxation factor λ to the different largest tolerable that produce of performance requirement.Be not more than remaining multi-user interference lax of largest tolerable with relaxation factor λ to disturbing alignment constraints, obtain disturbing the constraints of lax alignment.

2b) selected target function.According to making the maximized optimization criterion of channel capacity, it is maximum as the target function of optimizing equation that the equivalent independent channel that selected all receiving terminals obtain is counted sum.

2c) set up the optimization equation of forming by constraints and target function:

Target function: $\max \underset{k\∈K}{\mathrm{\Σ}}{d}_k$

Constraints: $\left\{\begin{array}{cc}{U}_k^H{H}_{\mathrm{kj}}{V}_j\≤{\mathrm{\σ}}_k^{2}I/\mathrm{\λ},& \∀k\≠j\\ \mathrm{rank}\left(U_k^H{H}_{\mathrm{kk}}{V}_k\right)=d_k,& 1\≤k\≤K\end{array}\right.$

Wherein, d _kBe the equivalent independent channel number of receiving terminal k acquisition,

U _kBe the interference cancellation matrix of k receiving terminal,

[] ^HBe the conjugate transpose of matrix,

H _KjBe the channel matrix parameter of j transmitting terminal to k receiving terminal,

V _jBe the beam shaping matrix of j transmitting terminal,

Be additive white Gaussian noise power,

I is a unit matrix,

Rank () is a rank of matrix,

H _KkBe the channel matrix parameter of k transmitting terminal to k receiving terminal,

V _kIt is the beam shaping matrix of k transmitting terminal.

(3) solving-optimizing equation obtains to be used to suppress the beam shaping set of matrices and the interference cancellation matrix set of multi-user interference, and sends to corresponding sending terminal and receiving terminal respectively.

(4) each transmitting terminal with beam shaping matrix corresponding in the beam shaping set of matrices that obtains and transmission signal premultiplication, is finished sending the preliminary treatment of signal respectively.

(5) signal that the conjugate transpose premultiplication of the interference cancellation matrix of correspondence received during each receiving terminal was gathered with the interference cancellation matrix that obtains is respectively finished reprocessing to the received signal.

The present invention compared with prior art has following advantage:

The first, the present invention has solved the puzzlement that the difficult problem of NP-can't be found the solution in the prior art owing to introduced relaxation factor, can go out to be used to suppress required beam shaping set of matrices of multi-user interference and interference cancellation matrix set by rapid solving.

Second, the present invention is not only applicable to transmitting terminal number and receiving terminal number all more than or equal to 3 situation, and be not all more than or equal to 3 o'clock at transmitting terminal number and receiving terminal number, also can use method rapid solving of the present invention and go out to be used to suppress required beam shaping set of matrices of multi-user interference and interference cancellation matrix set.

The 3rd, the present invention is based on the beam shaping set of matrices and the interference cancellation matrix set that solve, can carry out preliminary treatment and reprocessing respectively at transmitting terminal and receiving terminal, realize distinguishing the lax alignment of interference of disturbing and expecting received signal, suppress the phase mutual interference between the multi-user.

Description of drawings

Fig. 1 is a flow chart of the present invention.

Fig. 2 is the multiple spot multi-user downlink transfer schematic diagram of cooperating.

Embodiment

It is following that the invention will be further described with reference to accompanying drawing.

At as shown in Figure 2 the multiple spot multi-user downlink transfer scene of cooperating, the process that the present invention suppresses multi-user interference as shown in Figure 1, the step of its realization is as follows:

Step 1 obtains the channel matrix parameter.According to communication system different to the difference of performance requirement and transmission means, transmitting terminal adopts corresponding channel estimation technique to finish channel estimating, obtains the channel matrix parameter.

Channel estimation technique can be divided into two classes: blind estimation and based on the estimation of reference signal.Different channel estimation techniques obtains different availability of frequency spectrum performances, when communication system needs higher availability of frequency spectrum performance, can adopt blind estimation, and blind estimation utilizes the information of transmission data inherence to realize channel estimating fully.Obviously, bandwidth has been saved in blind estimation, has improved the availability of frequency spectrum of system, but its algorithm operation quantity is big, flexibility is relatively poor, has therefore also brought very big processing delay.On the contrary, when communication system needs better delay performance, can adopt channel estimating based on reference signal.Channel estimating based on reference signal is determined parameter to be estimated by certain estimation criterion, perhaps progressively follows the tracks of and adjust the estimated value of waiting to estimate parameter by some criterion.Being characterized in need be by reference signal, i.e. pilot tone or training sequence.Be referred to as algorithm for estimating based on the estimation of training sequence and pilot frequency sequence based on reference signal.Be applicable to the system of burst transfer mode based on the channel estimation method of training sequence.By sending known training sequence, carry out initial channel estimating at receiving terminal, when sending the Useful Information data, utilize initial channel estimation results to carry out a judgement and upgrade, finish real-time channel estimating.Be applicable to the system of continuous transmission means based on the channel estimating of frequency pilot sign.By in the useful data that sends, inserting known frequency pilot sign, can obtain the channel estimation results of pilot frequency locations; The channel estimation results that then utilizes pilot frequency locations obtains the channel estimation results of useful data position by interpolation, finishes channel estimating.The channel estimation technique of embodiments of the invention adopts least mean-square error (MMSE) channel estimation technique based on pilot tone, though the low delay performance of this technology availability of frequency spectrum is better.

MMSE channel estimation methods based on pilot tone is to insert pilot tone in the transmitting terminal appropriate location, and receiving terminal utilizes pilot tone to recover the channel information of pilot frequency locations, obtains all channel informations by means such as interpolation, filtering, conversion then.In the practice, the inserted mode of pilot tone has two kinds: Comb Pilot and block pilot tone.According to the difference of channel condition, select different pilot frequency designs to be used for channel estimating.Comb Pilot is that the channel estimating of carrying out with Comb Pilot is continuous on time domain in frequency domain interval ground distribution pilot tone; Block pilot tone is that the channel estimating of carrying out with block pilot tone is continuous on frequency domain in the time domain pilot tone that periodically distributes.Not under the Change channel condition, the performance of two kinds of pilot frequency designs is duplicate when additive white Gaussian noise (AWGN), but in Quick-Change channel, Comb Pilot is better than block pilot tone.Selectivity can better Comb Pilot be carried out channel estimating in the embodiments of the invention.

After receiving the symbol that contains Comb Pilot information, receiving terminal will recover pilot tone channel information constantly from pilot frequency locations according to the MMSE criterion.

The time domain MMSE estimated result that can be obtained channel by MMSE channel estimating formula is:

${\hat{h}}_{\mathrm{MMSE}}=R_{\mathrm{hY}}{R}_{\mathrm{YY}}^{-1}Y$

Wherein,

Be time domain MMSE estimated result,

R _HYBe the cross-correlation matrix of h and Y, h is a channel impulse response,

[] ^-1Be inverse matrix,

R _YYBe the autocorrelation matrix of Y,

Y is the frequency domain received signal that obtains after the receiving terminal process N point discrete Fourier conversion (DFT),

N is a discrete Fourier transform (DFT) length of an interval degree.

Based on above time domain estimated result, will

The frequency domain MMSE estimated result that multiplies each other and can obtain channel with N point DFT transformation matrix F:

Wherein,

$W_N^{\mathrm{nk}}=\frac{1}{\sqrt{N}}{e}^{-j2\mathrm{\π}\frac{\mathrm{nk}}{N}}.$

Step 2 is set up and is optimized equation.

2a), determine constraints according to the remaining multi-user interference of each system to the different largest tolerable that produce of performance requirement.

Send signal after process transmitting terminal preliminary treatment and Channel Transmission, the signal that arrives k receiving terminal is:

y _k＝H _kkV _ks _k+∑H _kjV _js _j+z _k

Wherein, y _kBe k the signal that receiving terminal is received,

H _KkBe the channel matrix parameter of k transmitting terminal to k receiving terminal,

V _kBe the beam shaping matrix of k transmitting terminal,

s _kBe the transmission signal of k transmitting terminal,

H _KjBe the channel matrix parameter of j transmitting terminal to k receiving terminal,

V _jBe the beam shaping matrix of j transmitting terminal,

s _jBe the transmission signal of j transmitting terminal,

z _kBe additive white Gaussian noise, Normal Distribution

Be noise power.The signal y that receiving terminal is received _kThrough reprocessing, the signal that finally demodulates is:

Wherein,

Be k the signal that receiving terminal demodulates,

U _kBe the interference cancellation matrix of k receiving terminal,

[] ^HThe conjugate transpose of representing matrix.

In second

H _KjV _jFor disturbing the remaining multi-user interference after aliging.The desirable constraints of alignment of disturbing is:

$U_k^H{H}_{\mathrm{kj}}{V}_j=0$ $(\∀k\≠j)$

As seen, disturb alignment constraints by the ideal that theoretical derivation obtains, require to realize disturbing the remaining multi-user interference in alignment back should be entirely zero, in practice because the existence of background noise, so desirable interference alignment there is no need fully, and the Project Realization of algorithm is too complicated.The remaining multi-user interference of the largest tolerable that each system produces the performance requirement difference is different, performance is meant the magnitude of remaining multi-user interference, determines that according to the remaining multi-user interference of the largest tolerable that different system produced corresponding relaxation factor λ should satisfy:

${\max }_{\mathrm{tol}}\left(U_k^H{H}_{\mathrm{kj}}{V}_j\right)={\mathrm{\σ}}_k^{2}I/\mathrm{\λ}$ $(\∀k\≠j)$

Wherein,

Be the remaining multi-user interference of largest tolerable,

I is a unit matrix.

Disturb alignment constraints to be not more than remaining multi-user interference lax of largest tolerable with the relaxation factor λ that determines to ideal, make the constraints of remaining multi-user interference below satisfying:

$U_k^H{H}_{\mathrm{kj}}{V}_j\≤{\mathrm{\σ}}_k^{2}I/\mathrm{\λ}$ $(\∀k\≠j)$

If the remaining multi-user interference of system requirements is smaller or equal to half of noise power, then λ=2.The constraints that remaining multi-user interference satisfies is:

$U_k^H{H}_{\mathrm{kj}}{V}_j\≤{\mathrm{\σ}}_k^{2}I/2$ $(\∀k\≠j)$

2b) selected optimization aim function.The equivalent independent channel number that receiving terminal can obtain is big more, and the channel capacity of its acquisition is also just big more.In order to make the channel capacity sum maximum of disturbing all receiving terminals of lax alignment back, according to making the maximized optimization criterion of channel capacity, optimization equation target function is chosen to be the equivalent independent channel that all receiving terminals are obtained counts the sum maximum, the optimization aim function expression of this moment is:

$\max \underset{k\∈K}{\mathrm{\Σ}}{d}_k,1\≤k\≤K$

Wherein, d _kBe the equivalent independent channel number of receiving terminal k acquisition,

Rank () is a rank of matrix.

2c) set up the optimization equation.Set up following by step 2a) in constrained conditions set and step 2b) in the optimization equation formed of selected target function:

Target function: $\max \underset{k\∈K}{\mathrm{\Σ}}{d}_k$

Constraints: $\left\{\begin{array}{cc}{U}_k^H{H}_{\mathrm{kj}}{V}_j\≤{\mathrm{\σ}}_k^{2}I/\mathrm{\λ},& \∀k\≠j\\ \mathrm{rank}\left(U_k^H{H}_{\mathrm{kk}}{V}_k\right)=d_k,& 1\≤k\≤K\end{array}\right.$

This optimizes equation is to satisfy under the constraints of the lax alignment remaining multi-user interference in back of interference smaller or equal to the remaining multi-user interference of system's largest tolerable, seeking the beam shaping set of matrices and the interference cancellation matrix that make channel capacity sum maximum and gather.

Step 3, the solving-optimizing equation obtains effectively separating of united beam shaping.By the optimization equation of setting up in the Combined Treatment center solution procedure 2 of transmitting terminal, obtain beam shaping set of matrices V _m(m=1 is 2...L) with interference cancellation matrix set U _n(n=1 2...K), and sends to corresponding sending terminal and receiving terminal respectively, and wherein the beam shaping set of matrices is used for the preliminary treatment of each transmitting terminal of step 4, and the interference cancellation matrix set is used for the reprocessing of each receiving terminal of step 5.

Step 4, the transmitting terminal preliminary treatment.Each transmitting terminal is respectively with the beam shaping set of matrices V that obtains in the step 3 _m(m=1,2...L) middle corresponding beam shaping matrix premultiplication sends signal and finishes pre-treatment step.

At transmitting terminal 1, use the beam shaping matrix V ₁Premultiplication sends signal s ₁, obtain V ₁s ₁, then with its transmission;

At transmitting terminal 2, use the beam shaping matrix V ₂Premultiplication sends signal s ₂, obtain V ₂s ₂, then with its transmission; At transmitting terminal L, use the beam shaping matrix V _LPremultiplication sends signal s _L, obtain V _Ls _L, then with its transmission.

Step 5, the receiving terminal reprocessing.Each receiving terminal is respectively with the interference cancellation matrix set U that obtains in the step 3 _n(n=1, the signal that the conjugate transpose premultiplication of corresponding interference cancellation matrix receives in 2...K) is finished post-processing step.

At receiving terminal 1, use interference cancellation matrix U ₁Conjugate transpose

The signal y that premultiplication receives ₁, the signal that is finally demodulated

At receiving terminal 2, use interference cancellation matrix U ₂Conjugate transpose

The signal y that premultiplication receives ₂, the signal that is finally demodulated

At receiving terminal K, use interference cancellation matrix U _KConjugate transpose

The signal y that premultiplication receives _K, the signal that is finally demodulated

Claims (6)

1. one kind based on the multiple spot multi-user lax alignment schemes of descending interference of cooperating, and comprises the steps:

(1) obtain the channel matrix parameter: according to communication system different to performance requirement and transmission means, transmitting terminal adopts corresponding channel estimation technique to finish channel estimating, obtains the channel matrix parameter;

(2) set up the optimization equation

2a) determine constraints, according to the remaining multi-user interference of each system to the different largest tolerable that produce of performance requirement, determine corresponding relaxation factor λ, be not more than remaining multi-user interference lax of largest tolerable with relaxation factor λ to disturbing alignment constraints, obtain disturbing the constraints of lax alignment;

2b) selected target function: according to making the maximized optimization criterion of channel capacity, it is maximum as the target function of optimizing equation that the equivalent independent channel that selected all receiving terminals obtain is counted sum;

2c) set up the optimization equation of forming by constraints and target function:

Target function: $\max \underset{k\∈K}{\mathrm{\Σ}}{d}_k$

Constraints: $\left\{\begin{array}{cc}{U}_k^H{H}_{\mathrm{kj}}{V}_j\≤{\mathrm{\σ}}_k^{2}I/\mathrm{\λ},& \∀k\≠j\\ \mathrm{rank}\left(U_k^H{H}_{\mathrm{kk}}{V}_k\right)=d_k,& 1\≤k\≤K\end{array}\right.$

Wherein, d _kBe the equivalent independent channel number of receiving terminal k acquisition,

U _kBe the interference cancellation matrix of k receiving terminal,

[] ^HBe the conjugate transpose of matrix,

H _KjBe the channel matrix parameter of j transmitting terminal to k receiving terminal,

V _jBe the beam shaping matrix of j transmitting terminal,

Be additive white Gaussian noise power,

I is a unit matrix,

Rank () is a rank of matrix,

H _KkBe the channel matrix parameter of k transmitting terminal to k receiving terminal,

V _kIt is the beam shaping matrix of k transmitting terminal;

(3) solving-optimizing equation obtains the set of beam shaping set of matrices and interference cancellation matrix, and sends to corresponding sending terminal and receiving terminal respectively;

(4) each transmitting terminal with beam shaping matrix corresponding in the beam shaping set of matrices that obtains and transmission signal premultiplication, is finished sending the preliminary treatment of signal respectively;

(5) signal that the conjugate transpose premultiplication of the interference cancellation matrix of correspondence received during each receiving terminal was gathered with the interference cancellation matrix that obtains is respectively finished reprocessing to the received signal.

2. according to claim 1 based on the multiple spot multi-user lax alignment schemes of descending interference of cooperating, it is characterized in that the performance of communication system comprises availability of frequency spectrum performance and delay performance in the described step (1).

3. according to claim 1ly it is characterized in that the communication system transmits mode comprises burst transfer mode and transmission means continuously in the described step (1) based on the multiple spot multi-user lax alignment schemes of descending interference of cooperating.

4. according to claim 1ly it is characterized in that the corresponding channel estimation technique in the described step (1) adopts blind estimation and based on the estimation technique of reference signal based on the multiple spot multi-user lax alignment schemes of descending interference of cooperating.

5. according to claim 1 based on the multiple spot multi-user lax alignment schemes of descending interference of cooperating, it is characterized in that performance is meant the magnitude of remaining multi-user interference in the described step (2).

6. according to claim 1 based on the multiple spot multi-user lax alignment schemes of descending interference of cooperating, it is characterized in that the corresponding relaxation factor λ that determines in the described step (2) should satisfy:

${\max }_{\mathrm{tol}}\left(U_k^H{H}_{\mathrm{kj}}{V}_j\right)={\mathrm{\σ}}_k^{2}I/\mathrm{\λ}$ $(\∀k\≠j)$

Wherein,

Be the remaining multi-user interference of largest tolerable,

U _kBe the interference cancellation matrix of k receiving terminal,

[] ^HThe conjugate transpose of representing matrix,

H _KjBe the channel matrix parameter of j transmitting terminal to k receiving terminal,

V _jBe the beam shaping matrix of j transmitting terminal,

Be additive white Gaussian noise power,

I is a unit matrix.

Images