CN102340473B - Noise signal estimation device and method - Google Patents

Abstract

The invention provides a noise signal estimation device and a method. The method comprises the steps of: selecting a subcarrier frequency band and constructing an interference matrix A (gamma) according to interference among subcarriers; and then calculating to obtain noise signals according to reference signals received by a receiving end and the interference matrix A (gamma). By adopting the technical scheme, the problem that noise is very difficult to estimate because of inter-carrier interference (ICI) caused by frequency deviation in the existing orthogonal frequency-division multiple access system is solved.

Description

A kind of estimation unit of noise signal and method

Technical field

The present invention relates to OFDMA (OFDM) system, be specifically related to a kind of estimation unit and method of noise signal.

Background technology

It is very important technology that noise is estimated in many communication systems, and the quality that noise is estimated can have influence on the performance of receiver.In OFDMA system, owing to there being the impact of Doppler frequency shift in crystal oscillator deviation and space, the frequency shift (FS) that can produce subcarrier, and cause thus inter-carrier interference (ICI), and these disturb and have affected to a certain extent the accuracy that noise is estimated.Therefore, receiving terminal with reference signal carry out noise estimate time because the existence of these interference can exist larger deviation.Current noise Estimation Algorithm is difficult to accurately estimate the noise level of current demand signal in the time that inter-carrier interference is larger.

Summary of the invention

Technical problem to be solved by this invention is to provide a kind of estimation unit and method of noise signal, can solve between the subcarrier that existing OFDMA system frequency offset causes and disturb (ICI) noise estimation value to be had to the problem of larger difficulty.

The method of estimation that the invention provides a kind of noise signal, comprising:

Choose a subcarrier frequency range, according to the interference structure interference matrix A (γ) between subcarrier; The reference signal and the described interference matrix A (γ) that receive according to receiving terminal afterwards calculate noise signal.

Further, in the subcarrier frequency range of choosing described in establishing, comprise n subcarrier;

In the interference matrix A (γ) of described structure, the interference of k the subcarrier of all the other subcarrier pairs of element representation that k is capable.

Further, the element of establishing described interference matrix is a _kl, represent the interference of k subcarrier of l subcarrier pair;

In the time of k=l, described in $a_{\mathrm{kl}}=\left\{\frac{\sin \left(\mathrm{\π\γ}\right)}{N\sin \left(\frac{\mathrm{\π\γ}}{N}\right)}\right\}\exp \left(\mathrm{j\π\γ}\frac{N-1}{N}\right);$

In the time of k ≠ l, described in $a_{k,l}=\left\{\frac{\sin \left(\mathrm{\π\γ}\right)}{N\sin \left(\frac{\mathrm{\π}(l-k+\mathrm{\γ})}{N}\right)}\right\}\exp (-\mathrm{j\π}\frac{l-k}{N})\exp \left(\mathrm{j\π\γ}\frac{N-1}{N}\right).$

Further, at the described interference matrix A of structure when (γ), only consider the interference of current subcarrier described near 3 subcarrier pairs current subcarrier, the current subcarrier of all the other subcarrier pairs is to disturb to be designated as 0.

Further, the described reference signal receiving according to receiving terminal and described interference matrix calculate noise signal and refer to,

Described A (γ) is disturbed to addition, then the sequence and the channel estimation sequence h that generate are multiplied each other, obtain one and add after interference the not reference signal sequence with noise, i.e. A (γ) * diag (reference) * h; If the reference signal that receiving terminal receives is receive_reference;

Described noise signal noise=receive_reference-A (γ) * diag (reference) * h;

Noise power spectral density ${\mathrm{\σ}}_{\mathrm{noise}}^2=C*\frac{1}{M}\underset{i=0}{\overset{M-1}{\mathrm{\Σ}}}{\left|{\mathrm{noise}}_i\right|}^2.$

The present invention also provides a kind of estimation unit of noise signal, comprises and chooses module and computing module;

The described module of choosing, for choosing a subcarrier frequency range;

Described conversion module, for constructing interference matrix A (γ) according to the interference between subcarrier;

Described computing module, calculates noise signal for the reference signal and the described interference matrix A (γ) that receive according to receiving terminal.

Further, in the subcarrier frequency range of choosing described in establishing, comprise n subcarrier;

In the interference matrix A (γ) of described conversion module structure, the interference of k the subcarrier of all the other subcarrier pairs of element representation that k is capable.

Further, the element of establishing described interference matrix A (γ) is a _kl, represent the interference of k subcarrier of l subcarrier pair;

In the time of k=l, described in $a_{\mathrm{kl}}=\left\{\frac{\sin \left(\mathrm{\π\γ}\right)}{N\sin \left(\frac{\mathrm{\π\γ}}{N}\right)}\right\}\exp \left(\mathrm{j\π\γ}\frac{N-1}{N}\right);$

In the time of k ≠ l, described in $a_{k,l}=\left\{\frac{\sin \left(\mathrm{\π\γ}\right)}{N\sin \left(\frac{\mathrm{\π}(l-k+\mathrm{\γ})}{N}\right)}\right\}\exp (-\mathrm{j\π}\frac{l-k}{N})\exp \left(\mathrm{j\π\γ}\frac{N-1}{N}\right).$

Further, described conversion module is at the described interference matrix A of structure when (γ), only considers the interference of current subcarrier described near 3 subcarrier pairs current subcarrier, and the current subcarrier of all the other subcarrier pairs is to disturb to be designated as 0.

Further, the reference signal that described computing module receives according to receiving terminal and described interference matrix calculate noise signal and refer to:

Described computing module disturbs addition to described A (γ), again the sequence and the channel estimation sequence h that generate are multiplied each other, obtain one and add after interference the not reference signal sequence with noise, i.e. A (γ) * diag (reference) * h; If the reference signal that receiving terminal receives is receive_reference;

Described computing module calculates noise signal as follows:

noise＝receive_reference-A(γ)*diag(reference)*h；

Corresponding noise power spectral density ${\mathrm{\σ}}_{\mathrm{noise}}^2=C*\frac{1}{M}\underset{i=0}{\overset{M-1}{\mathrm{\Σ}}}{\left|{\mathrm{noise}}_i\right|}^2.$

In sum, adopt the present invention can solve the problem that affects of disturbing (ICI) to estimate noise between the subcarrier that OFDMA system frequency offset causes, can estimate exactly the noise power-value under various channel circumstances according to the present invention.

Brief description of the drawings

Fig. 1 is apparatus of the present invention schematic diagrames;

Fig. 2 is the inventive method implementing procedure figure;

Fig. 3 be in LTE system sign bit and subcarrier be related to schematic diagram;

Fig. 4 is the interference value schematic diagram on k subcarrier and adjacent subcarrier in reference symbol.

Embodiment

The present embodiment provides a kind of estimation unit of noise signal, as shown in Figure 1, comprises and chooses module, computing module and conversion module;

Choose module, for choosing a subcarrier frequency range, establish this subcarrier frequency range and comprise n subcarrier;

Computing module, for estimating frequency offseting value γ according to frequency deviation algorithm for estimating, γ is normalized frequency offseting value;

γ=freq_offset/ Δ f, Δ f is the frequency interval between subcarrier, freq_offset obtains Frequency offset estimation value, the same prior art of evaluation method of freq_offset according to corresponding frequency offset estimation algorithm.

Computing module, also for local reference signal reference is disturbed to addition according to the interference matrix A (γ) obtaining, be A (γ) * diag (reference), again the sequence and the channel estimation sequence h that generate are multiplied each other, obtain one and add after interference the not reference signal sequence with noise, be A (γ) * diag (reference) * h, and use formula

Noise=receive_reference-A (γ) * diag (reference) * h calculates signal noise noise.

Computing module can also be used for using following formula calculating noise power spectral density:

${\mathrm{\σ}}_{\mathrm{noise}}^2=C*\frac{1}{M}\underset{i=0}{\overset{M-1}{\mathrm{\Σ}}}{\left|{\mathrm{noise}}_i\right|}^2.$

Conversion module, obtains interference matrix A (γ) element a for the expression formula of inter-carrier interference I (k) is converted _klexpression formula;

Wherein, k represents the row of interference matrix A (γ), and l represents the row of interference matrix A (γ), a _klrepresent the interference of k subcarrier of l subcarrier pair.

Preferably, conversion module, in the time of the described interference matrix A of structure (γ), is only considered near the interference of 3 current subcarriers of subcarrier pair of current subcarrier, and the current subcarrier of all the other subcarrier pairs is to disturb to be designated as 0.

The present embodiment provides a kind of method of estimation of noise signal, as shown in Figure 2, comprises the following steps:

Step S1: choose a subcarrier frequency range, establish this subcarrier frequency range and comprise n subcarrier;

Step S2: estimate frequency offseting value γ according to frequency deviation algorithm for estimating, γ is normalized frequency offseting value;

γ=freq_offset/ Δ f, Δ f is the frequency interval between subcarrier, freq_offset obtains Frequency offset estimation value, the same prior art of evaluation method of freq_offset according to corresponding frequency offset estimation algorithm.

Step S3: following ofdm signal formula (1) is done to corresponding conversion according to OFDM characteristic

$Y\left(k\right)=\left(X\left(k\right)H\left(k\right)\right)\left\{\frac{\sin \left(\mathrm{\π\γ}\right)}{N\sin \left(\frac{\mathrm{\π\γ}}{N}\right)}\right\}\exp \left(\mathrm{j\π\γ}\frac{N-1}{N}\right)+I\left(k\right)+W\left(k\right)---\left(1\right)$

Can derive the expression formula (2) of interference matrix and signal noise relation:

A(γ)*diag(reference)*h+noise＝receive_reference (2)

K is k subcarrier, and γ is normalized frequency offseting value, and I (k) disturbs (ICI) between subcarrier, and W (k) is noise sequence, and N is signal sampling point.

A (γ) is interference matrix, noise is signal noise, receive_reference is the reference signal that receiving terminal receives, reference is local reference signal, diag (reference) is the diagonal matrix of being write as according to local reference signal, h is channel estimation sequence, can calculate the channel estimation value h in current subcarrier frequency range according to the channel estimation method in OFDMA system.

And according to inter-carrier interference I (k) structure interference matrix A (γ);

Particularly, the expression formula of inter-carrier interference I (k) is:

$I\left(k\right)=\underset{l=0,l\≠k}{\overset{N}{\mathrm{\Σ}}}\left(X\left(l\right)H\left(l\right)\right)\left\{\frac{\sin \left(\mathrm{\π\γ}\right)}{N\sin \left(\frac{\mathrm{\π}(l-k+\mathrm{\γ})}{N}\right)}\right\}\exp \left(\mathrm{j\π\γ}\frac{N-1}{N}\right)\exp (-\mathrm{j\π}\frac{l-k}{N})---\left(3\right)$

I (k) represents the interference sum of all the other subcarrier pair subcarrier k;

Each element expression that can obtain in interference matrix A (γ) according to the expression formula of above-mentioned I (k) is:

In formula (4), k represents the row of interference matrix A (γ), and l represents the row of interference matrix A (γ), a _klrepresent the interference of k subcarrier of l subcarrier pair.

Formula (4) further can be used formula (5) and formula (6) to represent,

$\mathrm{\β}=\sin (\mathrm{\π}*\mathrm{\γ})*\exp (\mathrm{j\π}*\mathrm{\γ}*\left(\frac{N-1}{N}\right))---\left(5\right)$

w _i＝exp(-jπ*(i)/N)/(N*sin(π*(i+γ)/N))*β (6)

Formula (6) represents k+i subcarrier d _k+ito k subcarrier d _kinterference.

Can obtain thus interference matrix A (γ):

$A\left(\mathrm{\γ}\right)=\left(\begin{array}{ccccc}{a}_{11}& a_{12}& a_{13}& ......& a_{1n}\\ a_{21}& a_{22}& a_{23}& ......& a_{2n}\\ & & & ..........& \\ a_{k1}& a_{k2}& a_{k3}& ......& a_{\mathrm{kn}}\\ & & & ..........& \\ a_{n1}& a_{n2}& a_{n3}& ......& a_{\mathrm{nn}}\end{array}\right)$

In above-mentioned matrix, the interference of k subcarrier of the each subcarrier pair of k line display.N is the definite sampling number of system bandwidth, supposes that current subcarrier is k, a _k1represent the interference of the 1st k subcarrier of subcarrier pair, a _klrepresent the interference of k subcarrier of l subcarrier pair.

Step S4: to formula (2)

A(γ)*diag(reference)*h+noise＝receive_reference

Convert the computing formula (7) that can obtain noise:

noise＝receive_reference-A(γ)*diag(reference)*h (7)

Noise signal obtains through following process:

First local reference signal reference is disturbed to addition according to the interference matrix A (γ) obtaining, be A (γ) * diag (reference), again the sequence and the channel estimation sequence h that generate are multiplied each other, obtain one and add after interference the not reference signal sequence with noise, i.e. A (γ) * diag (reference) * h.

Receiving terminal extracts reference signal receive_reference after by OFDM demodulation, then deducts and add after interference not burst A (γ) * diag (reference) the * h with noise, is signal noise noise.

Afterwards, calculating noise power spectral density

${\mathrm{\σ}}_{\mathrm{noise}}^2=C*\frac{1}{M}\underset{i=0}{\overset{M-1}{\mathrm{\Σ}}}{\left|{\mathrm{noise}}_i\right|}^2,$ The noise power on each subcarrier is averaging.

Due to different according to the sub-carrier number of system bandwidth and user assignment, the noise power calculating need to be adjusted a constant coefficient, and C is a constant of determining according to the subcarrier number of distributing.

Application example

Be described in detail as an example of the subcarrier in LTE system example below.

In LTE system, a subframe has two time slots, has 12 subcarriers on a Resource Block RB, and subcarrier spacing is Δ f=15000Hz.The system bandwidth of supposing current employing is 20MHz, corresponding sampling number N=2048.The reference signal that transmitting terminal generates according to the rules condition is mapped in resource grid, and generates SC-FDMA symbol and launch.After channel, receiving terminal receives the data of eating dishes without rice or wine, and extracts the reference signal after channel, as shown in Figure 3.

The estimation procedure of noise is specific as follows:

Step 101: the reference signal receive_reference obtaining according to reception carries out LS channel estimating with local with reference to reference, takes out the sub-carrier signal on the milder RB of channel.

Step 102: adopt the channel estimation method in ofdm system to estimate the channel estimation value h on current RB, the same prior art of its method.

Step 103: estimated the frequency offseting value γ of subcarrier after channel by frequency deviation algorithm for estimating, Δ f=15000Hz.Suppose the frequency offseting value freq_offset=300Hz estimating, γ=freq_offset/ Δ f=0.02 after normalization.

Step 104: according to the frequency offseting value γ obtaining, calculate the interference matrix A (γ) on current RB.Be that adjacent nearest several subcarrier impacts are larger because most of ICI disturb, get near their impact of each 3 subcarriers calculating of the left and right of subcarrier so current, as shown in Figure 4.

Now N=2048

$\mathrm{\β}=\sin (\mathrm{\π}*0.02)*\exp (\mathrm{j\π}*0.02*\left(\frac{2048-1}{2048}\right))$

w _-3＝exp(-jπ*(-3)/2048)/(2048*sin(π*(-3+0.02)/2048))*β

w _-2＝exp(-jπ*(-2)/2048)/(2048*sin(π*(-2+0.02)/2048))*β

w _-1＝exp(-jπ*(-1)/2048)/(2048*sin(π*(-1+0.02)/2048))*β

w ₀＝exp(-jπ*(0)/2048)/(2048*sin(π*(0+0.02)/2048))*β

w ₁＝exp(-jπ*(1)/2048)/(2048*sin(π*(1+0.02)/2048))*β

w ₂＝exp(-jπ*(2)/2048)/(2048*sin(π*(2+0.02)/2048))*β

w ₃＝exp(-jπ*(3)/2048)/(2048*sin(π*(3+0.02)/2048))*β

$A\left(\mathrm{\γ}\right)=\left(\begin{array}{cccccccccccc}{w}_0& w_1& w_2& w_3& 0& 0& 0& 0& 0& 0& 0& 0\\ w_{-1}& w_0& w_1& w_2& w_3& 0& 0& 0& 0& 0& 0& 0\\ w_{-2}& w_{-1}& w_0& w_1& w_2& w_3& 0& 0& 0& 0& 0& 0\\ w_{-3}& w_{-2}& w_{-1}& w_0& w_1& w_2& w_3& 0& 0& 0& 0& 0\\ 0& w_{-3}& w_{-2}& w_{-1}& w_0& w_1& w_2& w_3& 0& 0& 0& 0\\ 0& 0& w_{-3}& w_{-2}& w_{-1}& w_0& w_1& w_2& w_3& 0& 0& 0\\ 0& 0& 0& w_{-3}& w_{-2}& w_{-1}& w_0& w_1& w_2& w_3& 0& 0\\ 0& 0& 0& 0& w_{-3}& w_{-2}& w_{-1}& w_0& w_1& w_2& w_3& 0\\ 0& 0& 0& 0& 0& w_{-3}& w_{-2}& w_{-1}& w_0& w_1& w_2& w_3\\ 0& 0& 0& 0& 0& 0& w_{-3}& w_{-2}& w_{-1}& w_0& w_1& w_2\\ 0& 0& 0& 0& 0& 0& 0& w_{-3}& w_{-2}& w_{-1}& w_0& w_1\\ 0& 0& 0& 0& 0& 0& 0& 0& w_{-3}& w_{-2}& w_{-1}& w_0\end{array}\right)$

Step 105: according to noise=receive_reference-A (γ) * diag (reference) * h calculating noise signal;

Local reference signal on current RB is disturbed to addition by matrix A (γ), then the sequence and the channel estimation sequence h that generate multiply each other, and obtain one and add after interference not reference signal sequence A (γ) * diag (reference) the * h with noise.Receiving terminal extracts the reference signal on current RB after by OFDM demodulation, then deducts and add after interference the not reference signal sequence with noise, just obtains the signal noise on current RB.

Step 106: calculating noise power spectral density

${\mathrm{\σ}}_{\mathrm{noise}}^2=C*\frac{1}{M}\underset{i=0}{\overset{M-1}{\mathrm{\Σ}}}{\left|{\mathrm{noise}}_i\right|}^2,$ Be that noise power on each subcarrier is averaging.

Due to different according to the sub-carrier number of system bandwidth and user assignment, the noise power calculating need to be adjusted a constant coefficient, and C is a constant of determining according to the subcarrier number of distributing.

One of ordinary skill in the art will appreciate that all or part of step in said method can carry out instruction related hardware by program and complete, described program can be stored in computer-readable recording medium, as read-only memory, disk or CD etc.Alternatively, all or part of step of above-described embodiment also can realize with one or more integrated circuits.Correspondingly, the each module/unit in above-described embodiment can adopt the form of hardware to realize, and also can adopt the form of software function module to realize.The present invention is not restricted to the combination of the hardware and software of any particular form.

Claims (8)

1. a method of estimation for noise signal, comprising:

Choose a subcarrier frequency range, according to the interference structure interference matrix A (γ) between subcarrier; The reference signal and the described interference matrix A (γ) that receive according to receiving terminal afterwards calculate noise signal;

The described reference signal receiving according to receiving terminal and described interference matrix calculate noise signal and refer to,

Described A (γ) is disturbed to addition, then the sequence and the channel estimation sequence h that generate are multiplied each other, obtain one and add after interference the not reference signal sequence with noise, i.e. A (γ) * diag (reference) * h; If the reference signal that receiving terminal receives is receive_reference;

Described noise signal noise=receive_reference-A (γ) * diag (reference) * h;

Noise power spectral density ${\mathrm{\σ}}_{\mathrm{noise}}^2=C*\frac{1}{M}\underset{i=0}{\overset{M-1}{\mathrm{\Σ}}}{\left|{\mathrm{noise}}_i\right|}^2.$

2. the method for claim 1, is characterized in that:

If described in comprise n subcarrier in the subcarrier frequency range chosen;

In the interference matrix A (γ) of described structure, the interference of k the subcarrier of all the other subcarrier pairs of element representation that k is capable.

3. method as claimed in claim 2, is characterized in that:

If the element of described interference matrix is a _kl, represent the interference of k subcarrier of l subcarrier pair;

In the time of k=l, described in $a_{\mathrm{kl}}=\left\{\frac{\sin \left(\mathrm{\π\γ}\right)}{N\sin \left(\frac{\mathrm{\π\γ}}{N}\right)}\right\}\exp \left(\mathrm{j\π\γ}\frac{N-1}{N}\right);$

In the time of k ≠ l, described in $a_{k,l}=\left\{\frac{\sin \left(\mathrm{\π\γ}\right)}{N\sin \left(\frac{\mathrm{\π}(l-k+\mathrm{\γ})}{N}\right)}\right\}\exp (-\mathrm{j\π}\frac{l-k}{N})\exp (-\mathrm{j\π\γ\γ}\frac{N-1}{N}).$

4. method as claimed in claim 2 or claim 3, is characterized in that:

At the described interference matrix A of structure when (γ), only consider the interference of current subcarrier described near 3 subcarrier pairs current subcarrier, the current subcarrier of all the other subcarrier pairs is to disturb to be designated as 0.

5. an estimation unit for noise signal, comprises and chooses module, conversion module and computing module; It is characterized in that:

The described module of choosing, for choosing a subcarrier frequency range;

Described conversion module, for constructing interference matrix A (γ) according to the interference between subcarrier;

Described computing module, calculates noise signal for the reference signal and the described interference matrix A (γ) that receive according to receiving terminal;

The reference signal that described computing module receives according to receiving terminal and described interference matrix calculate noise signal and refer to:

Described computing module disturbs addition to described A (γ), again the sequence and the channel estimation sequence h that generate are multiplied each other, obtain one and add after interference the not reference signal sequence with noise, i.e. A (γ) * diag (reference) * h; If the reference signal that receiving terminal receives is receive_reference;

Described computing module calculates noise signal as follows:

noise=receive_reference-A(γ)*diag(reference)*h；

Corresponding noise power spectral density ${\mathrm{\σ}}_{\mathrm{noise}}^2=C*\frac{1}{M}\underset{i=0}{\overset{M-1}{\mathrm{\Σ}}}{\left|{\mathrm{noise}}_i\right|}^2.$

6. device as claimed in claim 5, is characterized in that:

If described in comprise n subcarrier in the subcarrier frequency range chosen;

In the interference matrix A (γ) of described conversion module structure, the interference of k the subcarrier of all the other subcarrier pairs of element representation that k is capable.

7. device as claimed in claim 6, is characterized in that:

If the element of described interference matrix A (γ) is a _kl, represent the interference of k subcarrier of l subcarrier pair;

In the time of k=l, described in $a_{\mathrm{kl}}=\left\{\frac{\sin \left(\mathrm{\π\γ}\right)}{N\sin \left(\frac{\mathrm{\π\γ}}{N}\right)}\right\}\exp \left(\mathrm{j\π\γ}\frac{N-1}{N}\right);$

In the time of k ≠ l, described in $a_{k,l}=\left\{\frac{\sin \left(\mathrm{\π\γ}\right)}{N\sin \left(\frac{\mathrm{\π}(l-k+\mathrm{\γ})}{N}\right)}\right\}\exp (-\mathrm{j\π}\frac{l-k}{N})\exp (-\mathrm{j\π\γ\γ}\frac{N-1}{N}).$

8. the device as described in claim 6 or 7, is characterized in that:

Described conversion module is at the described interference matrix A of structure when (γ), only considers the interference of current subcarrier described near 3 subcarrier pairs current subcarrier, and the current subcarrier of all the other subcarrier pairs is to disturb to be designated as 0.

Images