CN101465828B - Integrity noise relevant estimation method and equipment - Google Patents

Abstract

The embodiment of the invention relates to a complete noise correlation estimation method and a complete noise correlation estimation device. The method includes the following steps: calculating the noise of all the antennas at the place i; calculating the correlation matrix of the noise; adding 1 to the value i, and continuing above process until the all noise of all the antennas at all the places and correlation matrixes of all the noise are calculated; wherein i is equal to 0,1,2,,,n-1(n is equal to the total of all the places). The device comprises a noise calculating unit, a matrix calculating module and a calculation indicating unit. The method and the device carry out noise vector calculation and matrix manipulation point by point, and only the current noise vector and the noise vector of the previous point are needed to be stored in each manipulation process. Therefore, the noise vectors of all the points are unnecessary to be stored before matrix manipulation in the process of storing the overall manipulation process, thus greatly reducing storage load and saving storage resource.

Description

Relevant method of estimation of complete noise and the relevant estimating apparatus of complete noise

Technical field

The present invention relates to communication technical field, relate in particular to relevant method of estimation of a kind of complete noise and the relevant estimating apparatus of complete noise.

Background technology

Global system for mobile communications (Global System for Mobile Communications, GSM) in, the signal that many antennas are collected need be merged, carry out equilibrium treatment.The mode that merges has three kinds: disturb and suppress to merge (Interference Rejection Combining, IRC), high specific merge (MaximumRatio Combining, MRC), interference cancellation (STIRC) when empty.

IRC, MRC and STIRC are based on the relevant estimation realization of complete noise.Wherein, IRC is to be based upon on the basis of the analysis of identical moment noise characteristic between antenna with MRC, and STIRC merging mode is based on the basis of analysis of identical and adjacent moment noise characteristic.

In the prior art, the relevant processing procedure of estimating of complete noise comprises:

One, calculates the noise of each all location point of antenna, promptly calculate the noise of being had a few.Wherein, location point is the sequence number of data just, and the data of this location point are corresponding with the time, also is the data in a certain moment.The data that transmit such as certain sequential of antenna are 8, and then the 1st data are first location point or data constantly, and the rest may be inferred.

At first, pass through formula:

$s_{\mathrm{nref}}\left(k\right)=(\mathrm{fix}\×(\underset{m=0}{\overset{L}{\mathrm{\Σ}}}{h}_n\left(m\right)\·\mathrm{TSC}(k-m)))$

Rebuild reference signal.

Wherein, k represents location point, k=L, and L+1 ... nTscLen-1, L represent the length of channel factors; N represents antenna amount, n=0, and 1 ... nAnts-1.

TSC in the following formula (k-m) is a training sequence, and (High Symbol Rate HSR) is 31 to high character rate, and (Low Symbol Rate LSR) is 26 to low character rate.Fix is the modifying factor of plural form, and obtaining value method is:

Under HSR, and 32 quadrature amplitude modulation (Quadrature Amplitude Modulation, in the time of QAM), fix=-2748-j4579; During 16QAM, fix=-3886-j3886; Quadrature Phase Shift Keying (QuadraturePhase-Shift Keying, in the time of QPSK), fix=-2896-j2896.

Under LSR, during 16QAM, fix=-3886-j3886; 8 phase shift keyings (Phase-shift keying, in the time of PSK), fix=4096; Gaussian-filtered minimum shift keying (Gaussian Filtered Minimum ShiftKeying, in the time of GMSK), fix=4096.

Under LSR MUROS, fix be for can join parameter, fix=α+i β, and α, β are configurable parameter. $\mathrm{fix}\×(\underset{m=0}{\overset{L}{\mathrm{\Σ}}}{h}_n\left(m\right)\·\mathrm{TSC}(k-m))$ It is complex multiplication.

Then, pass through formula:

u _n(k-L)＝r _n(k+nTaiLen+nDataLen)-s _nref(k)

Calculate the noise of each location point on all antennas.

Wherein, k=L, L+1 ... nTscLen-1, n=0,1 ... nAnts-1.

Two, calculate all noise correlation matrixs.

Pass through formula:

$\mathrm{corr}\left(n\right)=\left|\begin{array}{c}\mathrm{Corr}{\left(n\right)}_{\mathrm{0,0}}...\mathrm{Corr}{\left(n\right)}_{0,\mathrm{nAnts}-1}\\ \mathrm{Corr}{\left(n\right)}_{\mathrm{nAnts}-\mathrm{1,0}}...\mathrm{Corr}{\left(n\right)}_{\mathrm{nAnts}-1,\mathrm{nAnts}-1}\end{array}\right|$

Calculate all noise correlation matrixs.

Wherein, each the element corr (n) in the autocorrelation matrix passes through formula:

${\mathrm{Corr}\left(n\right)}_{i,j}=\underset{m=0}{\overset{\mathrm{nTscLen}-1-L-n}{\mathrm{\Σ}}}\left(u_i(m+n)u_j^{*}\left(m\right)\right)$

Calculate and obtain.Be about to u _i(m+n) u _j ^*(m) add up after taking advantage of again.

Corr32 (0) is that (be called for short: the while noise correlation matrix), its leading diagonal element is a real number, so imaginary part directly puts 0 with noise-plus-interference autocorrelation matrix constantly; Corr32 (1) is the noise plus interference correlation matrix (abbreviation: interim noise correlation matrix) in a moment that staggers.For the accuracy of fixing a point, omitted divided by (nTscLen-L) and calculating (nTscLen-1-L), this does not produce any influence to performance.

At this moment, finished the relevant processing of estimating of complete noise.Wherein, IRC and MRC are based on that Corr32 (0) analyzes, and STIRC is based on Corr32 (0), Corr32 (1) analyzes.

In realizing process of the present invention, the inventor finds that there is following defective at least in prior art.After the noise calculation of all antennas finished, carry out the matrix accumulation calculating again, cause the intermediate treatment process need to store the noise data of each antenna, and the size of memory space is relevant with antenna number with iterative times, iterative times is many more, antenna number is many more, and memory space is also big more.

Summary of the invention

The embodiment of the invention proposes relevant method of estimation of a kind of complete noise and the relevant estimating apparatus of complete noise, to reduce shared storage resources in the relevant estimation of the complete noise processing procedure.

The embodiment of the invention provides a kind of complete noise method of estimation of being correlated with, and comprising:

Calculating location i goes up the noise of all antennas;

Calculate the correlation matrix of described noise;

The i value is added 1, continue to carry out above-mentioned steps, until the correlation matrix of noise that calculates all antennas on all positions and noise; I=0 wherein, 1,2 ..., the quantity of all positions-1;

Wherein, the correlation matrix that calculates described noise comprises: noise correlation matrix or calculate the interim noise correlation matrix of described noise when calculating described noise.

The embodiment of the invention also provides a kind of complete noise estimating apparatus of being correlated with, and comprising:

The noise calculation unit is used for the noise that calculating location i goes up all antennas; I=0 wherein, 1,2 ..., the quantity of all positions-1;

Matrix calculation unit is used to calculate the correlation matrix of described noise;

Calculate indicating member, be used for that the described noise calculation of indication unit calculates under the situation of the quantity-1 of i≤all positions;

Wherein, described matrix calculation unit comprises:

Matrix unit simultaneously, noise correlation matrix when being used to calculate described noise;

The provisional matrix unit is used to calculate the interim noise correlation matrix of described noise.

The foregoing description carries out noise vector by pointwise and calculates and matrix operation, make each computing only need store current noise vector and more preceding noise vector, avoided carrying out storing the noise vector of being had a few before the matrix operation in the storage integral operation processing procedure, significantly reduce memory space, saved storage resources.

Below by drawings and Examples, technical scheme of the present invention is described in further detail.

Description of drawings

In order to be illustrated more clearly in the embodiment of the invention or technical scheme of the prior art, to do one to the accompanying drawing of required use in embodiment or the description of the Prior Art below introduces simply, apparently, accompanying drawing in describing below only is some embodiments of the present invention, for those of ordinary skills, under the prerequisite of not paying creative work, can also obtain other accompanying drawing according to these accompanying drawings.

Fig. 1 is the flow chart of the relevant method of estimation embodiment of complete noise of the present invention;

Fig. 2 is the logic realization structural representation of the relevant method of estimation embodiment of complete noise of the present invention;

Fig. 3 is the structural representation of the relevant estimating apparatus embodiment of complete noise of the present invention.

Embodiment

Below in conjunction with the accompanying drawing in the embodiment of the invention, the technical scheme in the embodiment of the invention is clearly and completely described, obviously, described embodiment only is the present invention's part embodiment, rather than whole embodiment.Based on the embodiment among the present invention, those of ordinary skills belong to the scope of protection of the invention not making the every other embodiment that is obtained under the creative work prerequisite.

Fig. 1 is the flow chart of the relevant method of estimation embodiment of complete noise of the present invention.Comprise:

Step 11, calculating location i go up the noise of all antennas, promptly calculate the noise of a point.

Step 12, utilize the correlation matrix of the noise of the noise calculation current location point that calculates in the step 11;

Step 13, the i value is added 1, continue to carry out above-mentioned steps 11, step 12, calculate the noise and the correlation matrix thereof of next point, until noise that calculates all antennas on all positions and noise correlation matrix; I=0 wherein, 1,2 ..., the quantity of all positions-1.

In the foregoing description, can be on calculating location i in the correlation matrix of noise, calculating location i+1 goes up the noise of all antennas.

In the above-mentioned steps 12, the correlation matrix that calculates the noise of current location point can comprise:

Noise correlation matrix when calculating the noise of current location point;

Calculate the interim noise correlation matrix of the noise of current location point.

In the present embodiment, carry out noise vector by pointwise and calculate and matrix operation, make each computing only need store the noise vector and the last location point of current location point

Noise vector, avoided carrying out storing the noise vector of being had a few before the matrix operation in the storage integral operation processing procedure, significantly reduced memory space, saved storage resources, simplified processing procedure, improve the relevant treatment effeciency of estimating of complete noise, thereby also saved the computing time that aerial signal merges, improved the treatment effeciency that merges.

The relevant processing procedure of estimating of complete noise with two antennas is an example, and pointwise computational process is described, that is, single calculation need be finished the calculating of noise vector and noise correlation matrix.

At first, calculate the noise of the same location point of each antenna, form noise vector:

$\stackrel{\‾}{u}\left(k\right)=\left[\begin{array}{c}{u}_1\left(k\right)\\ u_2\left(k\right)\end{array}\right].$

Wherein, u _i(k) computing formula is:

u _n(k-L)＝r _n(k+nTscLen+nDataLen)-s _nref(k)。

r _n(k+nTscLen+nDataLen) aerial signal on the expression position k+nTscLen+nDataLen.Wherein, TscLen represents the length of training sequence, and DataLen represents the length of valid data, also is the length of the user profile in the antenna data.

Then, calculate the correlation matrix of this spot noise.Comprise and calculate noise correlation matrix and interim noise matrix simultaneously.Wherein, simultaneously noise correlation matrix is promptly with noise-plus-interference autocorrelation matrix constantly, the interim noise matrix noise plus interference correlation matrix in a moment that promptly staggers.

Noise correlation matrix is simultaneously:

corr(0) _k＝u(k)u ^H(k)+corr(0) _k-1。

Interim noise correlation matrix is:

corr(1) _k＝u(k)u ^H(k-1)+corr(1) _k-1。

The logic realization structure of said method embodiment as shown in Figure 2, the noise vector computing module calculates the noise vector of current location point, and is buffered in the noise vector cache module.The noise vector U (k) that while correlation matrix module utilizes noise vector computing module compute to obtain obtains its orthogonal vectors U _HAnd utilize formula U (K) U (k), _H(k) calculate correlation matrix simultaneously.Interim correlation matrix module is further obtained the noise vector U (k-1) of last location point from the noise vector cache module, calculate to obtain its orthogonal vectors U _HAnd utilize formula U (K) U (k-1), _H(k-1) calculate interim correlation matrix.

Can rely on signed magnitude arithmetic(al) to realize for noise vector, promptly adopt adder directly to realize adding up, the calculating flowing water time of an antenna noise only needs 1 cycle (cycle) on the single-point like this, and so two calculating 2 cycles of needs that antenna is finished noise vector, four antennas need 4 cycles.

Then adopt the serial implementation for matrix operation, like this, finishing two antenna array computings needs 4 cycles, and finishing four antenna array computings needs 16 cycles.

According to above analysis as can be known: flowing water computing time of step 12 is always greater than flowing water computing time of step 11, increase the buffer memory of a noise vector in the middle of the logic, step 11 and step 12 are carried out synchronously, realize the time unification on two steps, promptly two antennas are 4 cycles, and four antennas are 16 cycles.Suppose that training sequence length is TscLen, the iterative times of then finishing noise vector and correlation matrix computing is inferior for (nTscLen-L), and wherein L is a channel length.Like this, the complete noise of a point of every realization is relevant estimates to handle the required time to be as the criterion computing time of step 12, finishes that the complete noise of being had a few is relevant to be estimated to multiply each other with regard to the time that equals iterative times and matrix operation computing time.And each matrix operation can obtain up-to-date noise vector in can guaranteeing to iterate.

Iterating 37 times with four antennas is that example and prior art compare, and said method embodiment has the following advantages:

Reduced the memory space of intermediate storage noise vector: 16 * 2 * 4 * 37=4736 bit (bit), wherein, 16 is the real part and the imaginary part bit wide of noise data, on behalf of data, 2 comprise real part and imaginary part, 4 representative antennas numbers, 37 represent iterative times;

Only be 16 * 37=592 cycle required computing time under same resource consumes, and with respect to required 16 * 37+4 computing time of prior art * 37=740 cycle, saved the processing time greatly, improved treatment effeciency;

Said method embodiment also supports the lasting evolution for algorithm, promptly for analyzing such as u (k) u (k-2), only needs to increase corresponding matrix operation unit and gets final product.

Fig. 3 is the structural representation of the relevant estimating apparatus embodiment of complete noise of the present invention.Equipment comprises: noise calculation unit 31, matrix calculation unit 32 and calculating indicating member 36.

Noise calculation unit 31 is used for the noise that calculating location i goes up all antennas; I=0 wherein, 1,2 ..., the quantity of all positions-1; Matrix calculation unit 32 is used to calculate the correlation matrix of described noise, and promptly matrix calculation unit 32 noise that utilizes noise calculation unit 31 to calculate obtains the correlation matrix of current location spot noise; Calculating indicating member 36 is used for indicating noise calculation unit 31 to calculate under the situation of the quantity-1 of i≤all positions.

Described matrix calculation unit 32 can comprise: while matrix unit 33, provisional matrix unit 34.Matrix unit 33 is used for calculating noise correlation matrix simultaneously simultaneously; Provisional matrix unit 34 is used to calculate interim noise correlation matrix.Correlation matrix when promptly simultaneously matrix unit 33 noise that utilizes noise calculation unit 31 to calculate obtains the current location spot noise; The noise that provisional matrix unit 34 utilizes noise calculation unit 31 to calculate obtains the interim correlation matrix of current location spot noise.

The relevant estimating apparatus of described complete noise also can comprise memory cell 35, and memory cell 35 is used to store the noise that noise calculation unit 31 calculates.

In said method and the apparatus embodiments,, carry out noise calculation based on pointwise by adjusting algorithm flow of the prior art, after the noise calculation of a point finishes, can finish matrix correlation computing, avoid the storage of middle noise vector, save storage resources for this point; And the concurrent operation of noise vector and matrix has also significantly reduced computing time.In addition, the relevant estimation of the complete noise that realizes in said method and the apparatus embodiments is applicable to that simultaneously three kinds of antennas merge mode

One of ordinary skill in the art will appreciate that: all or part of step that realizes said method embodiment can be finished by the relevant hardware of program command, aforesaid program can be stored in the computer read/write memory medium, this program is carried out the step that comprises said method embodiment when carrying out; And aforesaid storage medium comprises: various media that can be program code stored such as ROM, RAM, magnetic disc or CD.

It should be noted that at last: above embodiment only in order to technical scheme of the present invention to be described, is not intended to limit; Although with reference to previous embodiment the present invention is had been described in detail, those of ordinary skill in the art is to be understood that: it still can be made amendment to the technical scheme that aforementioned each embodiment put down in writing, and perhaps part technical characterictic wherein is equal to replacement; And these modifications or replacement do not make the essence of appropriate technical solution break away from the spirit and scope of various embodiments of the present invention technical scheme.

Claims (5)

1. the relevant method of estimation of complete noise is characterized in that, comprising:

Calculating location i goes up the noise of all antennas;

Calculate the correlation matrix of described noise;

The i value is added 1, continue to carry out above-mentioned steps, until the correlation matrix of noise that calculates all antennas on all positions and noise; I=0 wherein, 1,2 ..., the quantity of all positions-1;

Wherein, the correlation matrix that calculates described noise comprises: noise correlation matrix or calculate the interim noise correlation matrix of described noise when calculating described noise.

2. the complete noise according to claim 1 method of estimation of being correlated with is characterized in that, in the correlation matrix of noise, calculating location i+1 goes up the noise of all antennas on calculating location i.

3. the complete noise according to claim 1 and 2 method of estimation of being correlated with is characterized in that, also comprises: store the noise that described position i goes up all antennas.

4. the relevant estimating apparatus of complete noise is characterized in that, comprising:

The noise calculation unit is used for the noise that calculating location i goes up all antennas; I=0 wherein, 1,2 ..., the quantity of all positions-1;

Matrix calculation unit is used to calculate the correlation matrix of described noise;

Calculate indicating member, be used for that the described noise calculation of indication unit calculates under the situation of the quantity-1 of i≤all positions;

Wherein, described matrix calculation unit comprises:

Matrix unit simultaneously, noise correlation matrix when being used to calculate described noise;

The provisional matrix unit is used to calculate the interim noise correlation matrix of described noise.

5. the complete noise according to claim 4 estimating apparatus of being correlated with is characterized in that, also comprises:

Memory cell is used to store the noise that described noise calculation unit calculates.

Images