CN101174854B - Noise estimation method and device thereof - Google Patents

Abstract

The present invention discloses a method and a device for estimating noise. The method comprises that the transferred route measurement under different states of the receive sequence is calculated and the minimum route measurement among all route measurements is chosen; the minimum route measurement is taken as the estimated power of the noise signal. The present invention utilizes all data in a burst to estimate noise, so the obtained noise estimated value is more accurate.

Description

Noise estimation method and device

Technical field

The present invention relates to the signal processing technology field, be specifically related to a kind of noise estimation method and device.

Background technology

Noise Estimation in a lot of communication systems as time division multiple access (TDMA, Time Division MultipleAccess) in the system is a very important technology, no matter in the calculating of carrier/interface ratio, disturb and comprise: co-channel interference (CCI, Co-Channel Interference), the adjacent (ACI that disturbs frequently, Adjacent ChannelInterference) and intersymbol interference (ISI, in the application on inhibition Inter-Symbol Interference) and part upper strata, all need to use Noise Estimation.The quality of Noise Estimation can have influence on receiver performance.

At present, in gsm system, utilize training sequence and channel estimation results in the burst sequence (burst) usually, the noise in the burst is estimated as the narrow band TDMA system.Burst is the unit of transfer of gsm system, and 8 burst form a frame, corresponding 1 user of every frame.Each burst has 156.25 symbols; the structure of burst as shown in Figure 1; wherein; the data of 116 symbols (DATA) are divided into two parts; insert the training sequence (TSC) of 26 symbols between two parts data; the both sides of data respectively add the tail bit (TB) of 3 symbols, add the protection bit (GB) of 8.25 symbols at last at the burst end.

Noise Estimation in the gsm system can be used formula (1) expression:

$n\left(k\right)=r\left(k\right)-\underset{l=1}{\overset{L}{\mathrm{\Σ}}}p(k-i)*h\left(l\right)---\left(1\right)$

Wherein, r (k) is a receiving sequence, and p (k) is the training sequence among the burst, and h (l) is the channel estimating factor, and n (k) is the noise signal of estimating, L is the channel dispersion length of transmitting terminal to receiving terminal.

As can be seen, only utilized 26 training sequences among the burst in the existing noise estimation method, only under the more stable situation of the interference in a burst, the noise that this method estimates just can compare accurately.Yet in systems such as GSM, in fact probably there is asynchronous interference, i.e. interference instability in a burst, at this moment the noise signal that will estimate according to training sequence is irrational as the noise estimation value of whole burst, finally can cause receiver performance to descend.

Summary of the invention

The embodiment of the invention provides a kind of noise estimation method and device, to improve the Noise Estimation accuracy.

The technical scheme of the embodiment of the invention is achieved in that

A kind of noise estimation method comprises:

Calculate the path metric of receiving sequence under different conditions shifts, in all path metrics, select minimal path tolerance, with the estimating power of this minimal path tolerance as noise signal.

A kind of noise estimation device comprises:

The path metric computing module according to the receiving sequence and the channel estimating factor, calculates the path metric of receiving sequence under different conditions shifts;

The Noise Estimation module according to the path metric value that described path metric computing module calculates, is selected minimal path tolerance, with the estimating power of this minimal path tolerance as noise signal.

Compared with prior art, in the embodiment of the invention,, in all path metrics, select minimal path tolerance, with the estimating power of this minimal path tolerance as noise signal by calculating the path metric of receiving sequence under different conditions shifts.The embodiment of the invention has used all interior data of burst to come estimating noise, and therefore, resulting noise estimation value is more accurate.

Description of drawings

Fig. 1 is the structure chart of the burst in the gsm system;

The flow chart of the Noise Estimation that Fig. 2 provides for the embodiment of the invention;

The composition schematic diagram of the noise estimation device that Fig. 3 provides for the embodiment of the invention.

Embodiment

The present invention is further described in more detail below in conjunction with drawings and the specific embodiments.

In GSM, Wideband Code Division Multiple Access (WCDMA) systems such as (WCDMA, Wideband Code Division MultipleAddress), cause that by multipath channel frequency selective fading can cause intersymbol interference.Balancing technique is a kind of effective antagonism intersymbol interference, the technology of compensate for channel frequency selectivity.Maximum-likelihood sequence estimation (MLSE, Maximum Likelihood Sequence Estimation) equilibrium is the extraordinary equalization methods of a kind of performance, and it is basic demodulation techniques.

Under additive white Gaussian noise (AWGN, Additive White Gaussian Noise) condition, the MLSE equilibrium is best on the meaning that makes sequence error probability minimum.

As shown in Figure 2, the noise estimation method that provides of the embodiment of the invention can be as follows:

201: receiving sequence r (k).

Suppose that the sequence that sends among the D is { I between the area of observation coverage _k), the signal that this moment, receiver was received is r (t).

202: carry out the MLSE equilibrium according to r (k) and channel estimating factor h (l), calculate path metric the J ({ I of receiving sequence r (k) under different conditions shifts _k), wherein, { I _kIt is the possible state set that sends sequence.

According to maximum-likelihood criterion, to { I _kOptimal estimation

Should make likelihood function P (r (t), t ∈ D|{I _k) maximum, that is:

$P(r\left(t\right),t\∈D|\left\{{\hat{I}}_k\right\})=\underset{\left\{I_k\right\}}{\max }\left\{P(r\left(t\right),t\∈D|\left\{{\hat{I}}_k\right\})\right\}---\left(2\right)$

Find out and satisfy formula (2)

Method be the MLSE equilibrium.

And, P (r (t), t ∈ D|{I _k) be proportional to $-{\∫}_0^D|r\left(t\right)-\underset{k}{\mathrm{\Σ}}{I}_{k}h(t-\mathrm{kT})|\mathrm{dt},$ Therefore as can be known: only require out feasible $-{\∫}_0^D|r\left(t\right)-\underset{k}{\mathrm{\Σ}}{I}_{k}h(t-\mathrm{kT})|\mathrm{dt}$ Maximum

Just make ${\∫}_0^D|r\left(t\right)-\underset{k}{\mathrm{\Σ}}{I}_{k}h(t-\mathrm{kT})|\mathrm{dt}$ Minimum Just can be met formula (2)

Can adopt the Viterbi algorithm to realize the MLSE equilibrium:

The core of Viterbi algorithm is to calculate receiving sequence r _kPath metric J ({ I under different conditions shifts _k):

$J\left(\right\{I_k\left\}\right)=\underset{k=1}{\overset{K}{\mathrm{\Σ}}}{|r_k-\underset{l=1}{\overset{L}{\mathrm{\Σ}}}{h}_l\left(I_{k-l}\right)|}^2---\left(3\right)$

Wherein, { I _kIt is the possible state set that sends sequence; K is the sequence length of input MLSE equalizer, in gsm system, and the length of the longest desirable burst; L is the channel dispersion length of transmitting terminal to receiving terminal.

203: at all J ({ I _k) in, select minimum J ({ I _k) promptly: minJ ({ I _k) as the estimating power p of noise signal _n(k).

Because:

$\left\{{\hat{I}}_k\right\}=\arg \underset{\left\{I_k\right\}}{\min }J\left(\right\{I_k\left\}\right)---\left(4\right)$

That is: make $J\left(\right\{I_k\left\}\right)=\underset{k=1}{\overset{K}{\mathrm{\Σ}}}{|r_k-\underset{l=1}{\overset{L}{\mathrm{\Σ}}}{h}_l\left(I_{k-l}\right)|}^2$ Minimum

For sending sequence

Optimal estimation, thereby can learn: make according to this $J\left(\right\{I_k\left\}\right)=\underset{k=1}{\overset{K}{\mathrm{\Σ}}}{|r_k-\underset{l=1}{\overset{L}{\mathrm{\Σ}}}{h}_l\left(I_{k-l}\right)|}^2$ Minimum

The noise signal that estimates is inevitable the most approaching with the actual noise signal.

Because the general expression formula of noise signal is:

$n\left(k\right)=r_k-\underset{l}{\mathrm{\Σ}}{h}_l{I}_{k-l}---\left(5\right)$

With formula (5) and formula (3) contrast, as can be known:

$p_n\left(k\right)=\arg \underset{J\left(\right\{I_k\left\}\right)}{\min }J\left(\right\{I_k\left\}\right)---\left(6\right)$

Wherein, p _n(k) be the power of noise signal.

The composition schematic diagram of the noise estimation device that Fig. 3 provides for the embodiment of the invention, as shown in Figure 3, it mainly comprises: path metric computing module 31 and Noise Estimation module 32, wherein:

Path metric computing module 31: carry out the MLSE equilibrium according to receiving sequence r (k) and channel estimating factor h (l), calculate path metric the J ({ I of receiving sequence under different conditions shifts _k);

Noise Estimation module 32: each path metric value the J ({ I that calculates according to path metric computing module 31 _k), at all path metric value J ({ I _k) middle path metric value the minJ ({ I that selects minimum _k), with this minimal path metric minJ ({ I _k) as the estimating power p of noise signal _n(k).

The noise estimation device that the embodiment of the invention provides can be positioned on the receiver.

By the embodiment of the invention as can be seen: the embodiment of the invention has been used all data that burst is interior, therefore the noise signal that estimates is more accurate, in burst, exist non-stationary to disturb as: under the situation of asynchronous interference, can improve the performance of receiver greatly.Because, utilizing the MLSE equalization algorithm to obtain sending the sequence optimal estimation

Process in, just can obtain the minimal path metric, thereby obtain the estimating power of noise signal, therefore, the embodiment of the invention does not increase the operand of receiver, realizes simple.

Any system that has adopted the MSLE equilibrium to estimate optimum transmission sequence, the method that all can use the embodiment of the invention to provide is carried out Noise Estimation.

The above only is process of the present invention and method embodiment, in order to restriction the present invention, all any modifications of being made within the spirit and principles in the present invention, is not equal to replacement, improvement etc., all should be included within protection scope of the present invention.

Claims (5)

1. a noise estimation method is characterized in that, this method comprises:

According to maximum-likelihood sequence estimation (MLSE) equilibrium, calculate the path metric of receiving sequence under different conditions shifts;

In described path metric, select minimal path tolerance, with the estimating power of this minimal path tolerance as noise signal.

2. the method for claim 1 is characterized in that, describedly comprises according to the balanced path metric of receiving sequence under different conditions shifts that calculate of MLSE:

$J\left(\right\{I_k\left\}\right)=\underset{k=1}{\overset{K}{\mathrm{\Σ}}}{|r_k-\underset{l=1}{\overset{L}{\mathrm{\Σ}}}{h}_l\left(I_{k-l}\right)|}^2$

Wherein, { I _kFor sending the possible state set of sequence, J ({ I _k) be that receiving sequence is based on { I _kPath metric under different conditions shifts, r _kBe receiving sequence, K is the sequence length of input MLSE equalizer, h _lBe the channel estimating factor, L is the channel dispersion length of transmitting terminal to receiving terminal, I _K-lFor sending the possible state of the k-l symbol during moment in the sequence.

3. method as claimed in claim 2 is characterized in that, described K is the length of a burst sequence.

4. a noise estimation device is characterized in that, this device comprises:

The path metric computing module carries out maximum-likelihood sequence estimation (MLSE) equilibrium according to the receiving sequence and the channel estimating factor, calculates the path metric of receiving sequence under different conditions shifts;

The Noise Estimation module according to the path metric that described path metric computing module calculates, is selected minimal path tolerance, with the estimating power of this minimal path tolerance as noise signal.

5. device as claimed in claim 4 is characterized in that this device is positioned on the receiver.

Images