CN103200138A - Noise variance estimation method - Google Patents

Abstract

The invention provides a noise variance estimation method. The noise variance estimation method comprises the steps of calculating an amplitude value of channel gain of a pilot signal in a target sub-carrier according to channel frequency response of the pilot signal in at least one target sub-carrier in orthogonal frequency division multiplexing (OFDM) signs, wherein the target sub-carrier is located in the OFDM signs and used for bearing the pilot signal; obtaining a peak value of the channel gain of the pilot signal in the OFDM signs according to the amplitude value of the channel gain of the pilot signal in at least one target sub-carrier; and obtaining estimated values of noise variances of all of signals in the OFDM signs according to the peak value of the channel gain of the pilot signal in the OFDM signs, secondary moments of all of signals in the OFDM signs and forth moments of all of signals in the OFDM signs. By considering the influence of channel conditions on the estimated values of the noise variances, deviation between the estimated values of the noise variances in receiving signals and theoretical values is small under the condition of high signal-to-noise ratio, and the estimating performance of the noise variances is effectively improved.

Description

The Noise Variance Estimation method

Technical field

The present invention relates to mobile communication technology and aviation wide-band mobile communication technology, relate in particular to a kind of Noise Variance Estimation method.

Background technology

Along with development of Communication Technique, aviation aircrafts such as airline carriers of passengers can utilize aviation wide-band mobile communication technology, carry out broadband data communication with the communication network on ground, and the passenger can use the terminal equipment online or converse etc. in passenger plane.Aviation wide-band mobile communication technology provides the key of air traffic control information service of new generation.

L-band numeral aviation communication system (L-Band Digital Aeronautical Communications System Type1, L-DACS1) be a kind of in the aviation wide-band mobile communication system, adopt OFDM (Orthogonal Frequency Division Multiplexing, OFDM) technology.In ofdm system, Noise Variance Estimation is to carry out least mean-square error (Minimum Mean Square Error, MMSE) basis of balanced and soft demodulation is to guarantee the ofdm system transmission quality, brings into play the key of ofdm system superiority accurately.

The method of Noise Variance Estimation of the prior art comprises two classes: a class is based on the auxiliary Noise Variance Estimation method of pilot data, for example based on least square method (Least Square, LS) the time domain method of estimation of channel estimating; Another kind of is blind estimating method, for example second moment Fourth-order moment (M2M4) order algorithm.

Characteristics based on the auxiliary Noise Variance Estimation method of pilot data are comparatively simple, are easy to realize, but only are applied in pilot configuration and pilot frequency sequence satisfies in the system of particular requirement, can obtain good performance.And L-DACS1 adopts the design of dressing pilot configuration, its pilot frequency sequence is sparse relatively and unique, do not satisfy based on the requirement of the auxiliary Noise Variance Estimation method of pilot data to pilot configuration and pilot frequency sequence, therefore, adopt such based on the auxiliary Noise Variance Estimation method of pilot data, when the noise variance among the L-DACS1 is estimated, the problem that the performance of existence is lower, therefore, the available technology adopting blind estimating method is estimated the noise variance among the L-DACS1.

When blind estimating method is carried out the estimation of noise variance for the signal that the white Gaussian noise channel is transmitted, owing in the white Gaussian noise channel noise is only arranged, and transmission path is single-pathway, and estimated value and the square mean error amount between the ideal value of the noise variance that obtains are less.That is to say that the accuracy of utilizing blind estimating method that the noise variance of white Gaussian noise is estimated is higher.

But the transmission path among the L-DACS1 between transmitter and receiver is mulitpath, not only comprise the signal that transmitter sends in the signal that receiver receives, also comprise the noise in the channel, and channel response changes constantly, that is to say that the channel among the L-DACS1 is the multipath time varying channel.Under such transmission conditions, if still adopt blind estimating method noise variance to the received signal to estimate, then when the signal to noise ratio that receives signal is higher, because noise power is lower, when noise variance is estimated, be subjected to the influence of changing value of channel response bigger, thereby make under the situation of high s/n ratio, the estimated value and the square mean error amount between the ideal value that utilize blind estimating method that noise variance is estimated are bigger, that is to say that the accuracy of utilizing blind estimating method that the noise variance of signal among the L-DACS1 is estimated is lower.

Summary of the invention

The invention provides a kind of Noise Variance Estimation method, be used for solving when utilizing blind estimating method that the noise variance of L-DACS1 signal is estimated the problem that accuracy is lower.

First aspect of the present invention provides a kind of Noise Variance Estimation method, comprising:

Channel frequency response according to pilot signal at least one target sub-carriers in the orthogonal frequency division multiplex OFDM symbol, calculate the amplitude of the channel gain of pilot signal in each target sub-carriers, described target sub-carriers is the subcarrier that is used for the carrying pilot signal in the described OFDM symbol;

According to the amplitude of the channel gain of pilot signal in described at least one target sub-carriers, obtain the peak value of the channel gain of pilot signal in the described OFDM symbol;

According to whole whole Fourth-order moment of signals in the second moment of signals and the described OFDM symbol in the peak value of the channel gain of pilot signal in the described OFDM symbol, the described OFDM symbol, obtain in the described OFDM symbol all estimated values of the noise variance of signals.

The Noise Variance Estimation method that the embodiment of the invention provides, utilize the channel frequency response that receives pilot signal in the signal, calculate the peak value of the channel gain of pilot signal in each OFDM symbol, and then according to the peak value of the channel gain of pilot signal in each OFDM symbol, second moment and the Fourth-order moment of whole signals in each OFDM symbol, calculate the estimated value of the noise variance of whole signals in each OFDM symbol, by carrying out in the estimation approach to received signal noise variance, considered the influence of channel condition to the Noise Variance Estimation value, even make under the condition of high s/n ratio, the deviation of the estimated value of middle noise variance and theoretical value also can be smaller to received signal, improved effectively noise variance is carried out estimation performance.

Description of drawings

The flow chart of the Noise Variance Estimation method that Fig. 1 provides for the embodiment of the invention.

Embodiment

The flow chart of the Noise Variance Estimation method that Fig. 1 provides for the embodiment of the invention, as shown in Figure 1, this method comprises:

101, according to the channel frequency response of pilot signal at least one target sub-carriers in the OFDM symbol, calculate the amplitude of the channel gain of pilot signal in each target sub-carriers.

Wherein, described target sub-carriers is the subcarrier that is used for the carrying pilot signal in the described OFDM symbol.

Concrete, Noise Variance Estimation method in the embodiment of the invention, when the signal that receiver is received carries out Noise Variance Estimation, not only need consider to receive signal itself, also need to consider to receive the channel frequency response of pilot signal in channel in the signal.

Receiver receives signal at the OFDM of appointment symbol, and every frame signal takies one or more OFDM symbols, and each OFDM symbol is divided into one or more subcarriers at frequency domain.That is to say that subcarrier is the least unit that carrying receives signal.

According to the communication protocol between the transmitter and receiver, in the frame structure of every frame signal, pilot signal and data-signal according to certain regular distribution, correspondingly, in the subcarrier of each OFDM symbol, comprise for the subcarrier of carrying pilot signal and be used for the subcarrier of carry data signals.In the embodiment of the invention, the subcarrier that wherein carries pilot signal is called target sub-carriers.

Before step 101, this method can comprise: at least one the OFDM symbol from the every frame signal that receives, obtain described at least one target sub-carriers in described at least one OFDM symbol respectively; Calculate the channel frequency response of pilot signal in each target sub-carriers in described at least one OFDM symbol.

Concrete, for the every frame signal in receiving signal, from the subcarrier that carries whole signals, extract one or more target sub-carriers of wherein carrying pilot signal.

When the target sub-carriers that extracts when being a plurality of, each target sub-carriers can belong to identical OFDM symbol respectively, also can belong to different OFDM symbols respectively.

For the one or more target sub-carriers that extract, calculate the channel frequency response of the pilot signal in each target sub-carriers respectively.

Correspondingly, after the estimated value of the noise variance of whole signals, described method can also comprise in the described OFDM symbol of described acquisition:

Utilize the described every frame signal that receives, and whole estimated value of the noise variance of signals at least one the OFDM symbol in every frame signal, restore the signal that transmitter is launched.

Concrete, calculating after the estimated value of the noise variance of whole signals in each OFDM symbol, the estimated value of noise variance can be used for the signal to noise ratio of estimated signal, can also provide parameter for carrying out signal equalization.

According to the Noise Variance Estimation value of received every frame signal and each the OFDM symbol correspondence in every frame signal, can be by signal to be carried out equilibrium, compensate for channel is to the influence of signal, thereby restores the signal that transmitter is launched.

Further, the channel frequency response of pilot signal is in the described target sub-carriers, the ratio between the amplitude theoretical value of the amplitude of the actual pilot signal that receives on the described target sub-carriers and described pilot signal.

Suppose that frame data take a plurality of OFDM symbols, wherein comprise a plurality of subcarriers in each OFDM symbol, some subcarrier is the target sub-carriers for the carrying pilot signal in whole subcarriers; The quantity of target sub-carriers is less than the sum of whole subcarriers.

Receiving terminal can know in advance that transmitter in the pilot signal that subcarrier sends, that is to say according to communication protocol before receiver receives signal, know the amplitude theoretical value Xp of the pilot signal of carrying in each target sub-carriers in advance at receiving terminal _{N, k}Wherein, Xp _{N, k}The amplitude theoretical value of representing the pilot signal on k the subcarrier in n the OFDM symbol.

The amplitude of the actual pilot signal that receives of receiving terminal is Yp _{N, k}, Yp _{N, k}The actual amplitude that receives the pilot signal on k the subcarrier of signal in n OFDM symbol of expression.

Correspondingly, the channel frequency response H of the pilot signal on k the subcarrier in n OFDM symbol _{N, k}For,

Thereby, can calculate the channel frequency response of the pilot signal on each target sub-carriers of each OFDM symbol.

In the whole subcarriers in n the OFDM symbol, be not all to be used for the carrying pilot signal, therefore, for non-target sub-carriers wherein, there is not the channel frequency response H of pilot signal _{N, k}For the channel frequency response of pilot signal in the clearer expression target sub-carriers, below the channel frequency response of pilot signal in the target sub-carriers is called H _{N, j}

Further, this method can also comprise: the mould value of calculating the channel frequency response of pilot signal in the described target sub-carriers; With the mould value of the channel frequency response of pilot signal in the described target sub-carriers, as the amplitude of the channel gain of pilot signal in the described target sub-carriers.

Concrete, the channel frequency response H of pilot signal in calculating target sub-carriers _{N, j}Afterwards, to H _{N, j}Carry out the computing of delivery value, obtain | H _{N, j}|, will | H _{N, j}| as the amplitude of the channel gain of pilot signal in j the target sub-carriers in n the OFDM symbol.

102, according to the amplitude of the channel gain of pilot signal in described at least one target sub-carriers, obtain the peak value of the channel gain of pilot signal in the described OFDM symbol.

Concrete, after the amplitude of the channel gain of pilot signal, calculate the peak value of the channel gain of pilot signal in each OFDM symbol further in each target sub-carriers in calculating each OFDM symbol.

Concrete computational methods are,

To j target sub-carriers in each OFDM symbol | H _{N, j}| carry out the biquadratic computing, again this result of calculation of each OFDM symbol is carried out the average computing, what obtain is each OFDM symbol value corresponding respectively, with these numerical value as the denominator in the subsequent calculations.

To j target sub-carriers in each OFDM symbol | H _{N, j}| carry out the biquadratic computing, again this result of calculation in each OFDM symbol is carried out the average computing, what obtain is each OFDM symbol value corresponding respectively, and then these numerical value are carried out the quadratic power computing respectively, with these numerical value of obtaining as the molecule in the subsequent calculations.

Through after the above-mentioned calculating, in each numerical value of preparing as denominator and molecule, will with each OFDM symbol respectively value corresponding be divided by, that obtain is the peak value k of the channel gain of pilot signal in each OFDM symbol _{H, n}Correspondingly, resulting result of calculation is that each OFDM symbol is the peak value of the channel gain of corresponding pilot signal respectively.

103, according to whole whole Fourth-order moment of signals in the second moment of signals and the described OFDM symbol in the peak value of the channel gain of pilot signal in the described OFDM symbol, the described OFDM symbol, obtain in the described OFDM symbol all estimated values of the noise variance of signals.

Whole estimated value N of the noise variance of signals in n OFDM symbol _nFor,

Wherein, M _{2, n}Be the second moment of whole signals in the described OFDM symbol, M _{4, n}Fourth-order moment for whole signals in the described OFDM symbol.

y _{N, k}Be the signal in k the subcarrier in whole subcarriers of n OFDM symbol,

Be that product to the signal in each subcarrier in n the OFDM symbol and conjugated signal carries out the average computing.

Be the quadratic power to the product of the signal in each subcarrier in n the OFDM symbol and conjugated signal, the value of averaging computing.

Concrete computational process is as follows, supposes that receiving signal is y _{N, k}=a _{N, k}H _{N, k}+ w _{N, k}, n represents n OFDM symbol in the frame data, k represents k subcarrier in each OFDM symbol, a _{N, k}Represent the signal that transmitter is launched on k the subcarrier in n the OFDM symbol, w _{N, k}Represent the noise when signal transmits on k the subcarrier in n the OFDM symbol in channel, H _{N, k}The channel frequency response of representing signal on k the subcarrier in n the OFDM symbol.

In the above-described embodiments in order to show that channel frequency response is the frequency response of pilot signal, so utilize H _{N, j}Expression in the actual calculation process, is represented for the ease of parameter, can utilize H _{N, k}The expression channel frequency response.Be understandable that, if k subcarrier in n OFDM symbol is the target sub-carriers for the carrying pilot signal, then H _{N, k}Value is arranged; If k subcarrier in n OFDM symbol is not the target sub-carriers for the carrying pilot signal, then H _{N, k}No value.

Receive the second moment M of signal _{2, n}With Fourth-order moment M _{4, n}Can be expressed as:

$M_{2,n}=E\left\{y_{n,k}{y}_{n,k}^{*}\right\};$ $M_{4,n}=E\left\{{\left(y_{n,k}{y}_{n,k}^{*}\right)}^2\right\}.$

Be y _{N, k}Conjugated signal,

Because a _{N, k}, H _{N, k}And w _{N, k}All are plural forms, so with y _{N, k}Conjugated signal be expressed as:

$y_{n,k}^{*}=a_{n,k}^{*}{H}_{n,k}^{*}+w_{n,k}^{*}.$

And, by the calculating of plural number as can be known: $a_{n,k}{a}_{n,k}^{*}={\left|a_{n,k}\right|}^2,$ $H_{n,k}{H}_{n,k}^{*}={\left|H_{n,k}\right|}^2,$ $w_{n,k}{w}_{n,k}^{*}={\left|w_{n,k}\right|}^2,$ So with y _{N, k}With

Substitution

Calculate the second moment of each OFDM symbol:

$M_{2,n}=E\left\{y_{n,k}{y}_{n,k}^{*}\right\}=E\left\{(a_{n,k}{H}_{n,k}+w_{n,k})(a_{n,k}^{*}{H}_{n,k}^{*}+w_{n,k}^{*})\right\}$ $=E\{a_{n,k}{a}_{n,k}^{*}{H}_{n,k}{H}_{n,k}^{*}+a_{n,k}{H}_{n,k}{w}_{n,k}^{*}+w_{n,k}{a}_{n,k}^{*}{H}_{n.k}^{*}+w_{n,k}{w}_{n,k}^{*}\}$

Will $a_{n,k}{a}_{n,k}^{*}={\left|a_{n,k}\right|}^2,$ $H_{n,k}{H}_{n,k}^{*}={\left|H_{n,k}\right|}^2,$ With $w_{n,k}{w}_{n,k}^{*}={\left|w_{n,k}\right|}^2$ The substitution following formula obtains through further abbreviation:

$M_{2,n}=E\{{\left|a_{n,k}\right|}^2{\left|H_{n,k}\right|}^2+a_{n,k}{H}_{n,k}{w}_{n,k}^{*}+a_{n,k}^{*}{H}_{n,k}^{*}{w}_{n,k}+{\left|w_{n,k}\right|}^2\}$

$=E\left\{{\left|a_{n,k}\right|}^2{\left|H_{n,k}\right|}^2\right\}+E\left\{a_{n,k}{H}_{n,k}{w}_{n,k}^{*}\right\}+E\left\{a_{n,k}^{*}{H}_{n,k}^{*}{w}_{n,k}\right\}+E\left\{{\left|w_{n,k}\right|}^2\right\}$

With y _{N, k}With

Substitution Calculate the Fourth-order moment of each OFDM symbol:

$M_{4,n}=E\left\{{\left(y_{n,k}{y}_{n,k}^{*}\right)}^2\right\}=E\left\{{\left((a_{n,k}{H}_{n,k}+w_{n,k})(a_{n,k}^{*}{H}_{n,k}^{*}+w_{n,k}^{*})\right)}^2\right\}$

$=E\left\{\left({(a_{n,k}{H}_{n,k}+w_{n,k})}^2{(a_{n,k}^{*}{H}_{n,k}^{*}+w_{n,k}^{*})}^2\right)\right\}$ $=E\left\{(a_{n,k}^2{H}_{n,k}^2+2a_{n,k}{H}_{n,k}{w}_{n,k}+w_{n,k}^2)(a_{n,k}^{*2}{H}_{n,k}^{*2}+2a_{n,k}^{*}{H}_{n,k}^{*}{w}_{n,k}^{*}+w_{n,k}^{*2})\right\}$

$=E\left\{\begin{array}{c}{a}_{n,k}^2{H}_{n,k}^2{a}_{n,k}^{*2}{H}_{n,k}^{*2}+2a_{n,k}^{*2}{H}_{n,k}^2{a}_{n,k}^{*}{H}_{n,k}^{*}{w}_{n,k}^{*}+a_{n,k}^2{H}_{n,k}^2{w}_{n,k}^{*2}+2a_{n,k}{H}_{n,k}{w}_{n,k}{a}_{n,k}^{*2}{H}_{n,k}^{*2}+\\ 4a_{n,k}{H}_{n,k}{w}_{n,k}{a}_{n,k}^{*}{H}_{n,k}^{*}{w}_{n,k}^{*}+2a_{n,k}{H}_{n,k}{w}_{n,k}{w}_{n,k}^{*2}+w_{n,k}^2{a}_{n,k}^{*2}{H}_{n,k}^{*2}+2w_{n,k}^2{a}_{n,k}^{*}{H}_{n,k}^{*}{w}_{n,k}^{*}+w_{n,k}^2{w}_{n,k}^{*2}\end{array}\right\}$

Will $a_{n,k}{a}_{n,k}^{*}={\left|a_{n,k}\right|}^2,$ $H_{n,k}{H}_{n,k}^{*}={\left|H_{n,k}\right|}^2$ With $w_{n,k}{w}_{n,k}^{*}={\left|w_{n,k}\right|}^2$ The substitution following formula obtains through further abbreviation:

$M_{4,n}=E\left\{\begin{array}{c}{\left|a_{n,k}\right|}^4{\left|H_{n,k}\right|}^4+2{\left|a_{n,k}\right|}^2{\left|H_{n,k}\right|}^2{a}_{n,k}{H}_{n,k}{w}_{n,k}^{*}+2{\left|a_{n,k}\right|}^2{\left|H_{n,k}\right|}^2{a}_{n.k}^{*}{w}_{n,k}+{\left(a_{n,k}{H}_{n,k}{w}_{n,k}^{*}\right)}^2+\\ 4{\left|a_{n,k}\right|}^2{\left|H_{n,k}\right|}^2{\left|w_{n,k}\right|}^2+{\left(a_{n,k}^{*}{H}_{n,k}^{*}{w}_{n,k}\right)}^2+2{\left|w_{n,k}\right|}^2{a}_{n,k}{H}_{n,k}{w}_{n,k}^{*}+2{\left|w_{n,k}\right|}^2{a}_{n,k}^{*}{H}_{n,k}^{*}{w}_{n,k}+{\left|w_{n,k}\right|}^4\end{array}\right\}$

$=E\left\{{\left|a_{n,k}\right|}^4{\left|H_{n,k}\right|}^4\right\}+2\left(E\right\{{\left|a_{n,k}\right|}^2{\left|H_{n,k}\right|}^2{a}_{n,k}{H}_{n,k}{w}_{n,k}^{*}\}+E\{{\left|a_{n,k}\right|}^2{\left|H_{n,k}\right|}^2{a}_{n,k}^{*}{w}_{n,k}\left\}\right)+$ $\left(E\right\{{\left(a_{n,k}{H}_{n,k}{w}_{n,k}^{*}\right)}^2\}+4E\{{\left|a_{n,k}\right|}^2{\left|H_{n,k}\right|}^2{\left|w_{n,k}\right|}^2\}+E\{{\left(a_{n,k}^{*}{H}_{n,k}^{*}{w}_{n,k}\right)}^2\left\}\right)+$

$2\left(E\right\{{\left|w_{n,k}\right|}^2{a}_{n,k}{H}_{n,k}{w}_{n,k}^{*}\}+E\{{\left|w_{n,k}\right|}^2{a}_{n,k}^{*}{H}_{n,k}^{*}{w}_{n,k}\left\}\right)+E\left\{{\left|w_{n,k}\right|}^4\right\}$

Because the mathematic formula of signal power and noise power is as follows:

$S_n=E\left\{a_{n,k}{a}_{n,k}^{*}\right\}=E\left\{{\left|a_{n,k}\right|}^2\right\}$ $N_n=E\left\{w_{n,k}{w}_{n,k}^{*}\right\}=E\left\{{\left|w_{n,k}\right|}^2\right\}$

$E\left\{a_{n,k}{a}_{n,k}^{*}{w}_{n,k}{w}_{n,k}^{*}\right\}=E\left\{a_{n,k}{a}_{n,k}^{*}\right\}E\left\{w_{n,k}{w}_{n,k}^{*}\right\}=S_n{N}_n$

Suppose that signal and noise are that separate average is 0 random process, namely the theorem of chain of events dynamic relationship is described, and a kind of state that develops is obeyed the process of probability distribution in other words conj.or perhaps, then can obtain following expression:

$E\left\{a_{n,k}{H}_{n,k}{w}_{n,k}^{*}\right\}=E\left\{a_{n,k}\right\}E\left\{H_{n,k}\right\}E\left\{w_{n,k}^{*}\right\}=0\×E\left\{H_{n,k}\right\}E\left\{w_{n,k}^{*}\right\}=0$

$E\left\{{\left|a_{n,k}\right|}^2{{|H}_{n,k}|}^2{a}_{n,k}{H}_{n,k}{w}_{n,k}^{*}\right\}=0$

$E\left\{{\left|a_{n,k}\right|}^2{{|H}_{n,k}|}^2{a}_{n,k}^{*}{H}_{n,k}^{*}{w}_{n,k}\right\}=0$

$E\left\{{\left|w_{n,k}\right|}^2{a}_{n,k}{H}_{n,k}{w}_{n,k}^{*}\right\}=0$

$E\left\{{\left|w_{n,k}\right|}^2{a}_{n,k}^{*}{H}_{n,k}^{*}{w}_{n,k}\right\}=0$

$E\left\{{\left(a_{n,k}{H}_{n,k}{w}_{n,k}^{*}\right)}^2\right\}=E\left\{a_{n,k}^2\right\}E\left\{H_{n,k}^2\right\}E\left\{{\left(w_{n,k}^{*}\right)}^2\right\}=0$

$E\left\{{\left(a_{n,k}{H}_{n,k}{w}_{n,k}^{*}\right)}^2\right\}+E\left\{{\left(a_{n,k}^{*}{H}_{n,k}^{*}{w}_{n,k}\right)}^2\right\}=0$

Through abbreviation, calculate M _{2, n}, obtain:

M _2,n=S _n+N _n

Order $k_a=\frac{E\left\{{\left|a_{n,k}\right|}^4\right\}}{E{\left\{{\left|a_{n,k}\right|}^2\right\}}^2}=\frac{E\left\{{\left|a_{n,k}\right|}^4\right\}}{{S_n}^2},k_h=\frac{E\left\{{\left|H_{n,k}\right|}^4\right\}}{E{\left\{{\left|H_{n,k}\right|}^2\right\}}^2},k_w=\frac{E\left\{{\left|w_{n,k}\right|}^4\right\}}{E{\left\{{\left|w_{n,k}\right|}^2\right\}}^2}=\frac{E\left\{{\left|w_{n,k}\right|}^4\right\}}{{N_n}^2},$ Then:

E{a _n,k| ⁴}=k _aS _n ²,E{w _n,k| ⁴}=k _wN _n ²

Calculate M _{4, n}, obtain:

M _4,n=k _ak _hS _n ²+4S _nN _n+K _wN _n ²

And then, obtain N _n=M _{2, n}-S _n

That is to say that noise power is that noise variance is as follows:

$N_n=M_{2,n}-\sqrt{\frac{-2M_{2,n}^2+M_{4,n}}{\frac{E\left\{{\left|H_{n,j}\right|}^4\right\}}{E{\left\{{\left|H_{n,j}\right|}^2\right\}}^2}-2}}.$

Signal is sent in the process of receiver through channel at transmitter, because transmission channel is multipath channel, signal can produce decay in various degree in transmission course, therefore receives the peak value k of the channel gain of signal _{H, n}Less than 1.

If k _{H, n}More big, S then _nMore little, correspondingly, the estimated value N of noise variance _nMore big; If k _{H, n}More little, S then _nMore big, correspondingly, the estimated value N of noise variance _nMore little.Because k _{H, n}Calculate according to the channel frequency response of pilot signal, what reflect is the situation of channel, therefore, when noise variance was estimated, estimated value can be subjected to the influence of channel situation, under different channel conditions, estimated result to noise variance is different, thereby by introducing the peak value of channel gain, the estimated value of noise variance is compensated, with according to different channel situation, obtain the estimated value of corresponding noise variance.

For the aviation mobile communications system, because the mobility of aircraft, the vacant lot transmission channel also presents the characteristic of multipath fading channel, and transmission channel has time-varying characteristics, and channel gain constantly changes.

Be under the channel condition of 3dB, 6dB, 9dB, 12dB, 15dB and 18dB in signal to noise ratio respectively, the Noise Variance Estimation method of utilizing blind estimating method and the embodiment of the invention to provide, noise variance is to received signal estimated.

When utilizing blind estimating method noise variance to received signal to estimate, when signal to noise ratio was more high, the estimated value of noise variance and the deviation of theoretical value were more big, and the fluctuation of the Noise Variance Estimation value between each frame data is bigger.For example, be under the situation of 18dB in signal to noise ratio, deviation and fluctuation are all very obvious.

When the Noise Variance Estimation method noise variance to received signal that utilizes the embodiment of the invention to provide was estimated, the estimated value of the noise variance of every frame data was whole mean value of the estimated value of the noise variance of signals in each OFDM symbol in these frame data.Owing in computational process, introduced the channel frequency response of pilot signal, when being estimated, plays noise variance the effect of compensation, therefore, even under the higher channel condition of signal to noise ratio, the estimated value of noise variance and the deviation of theoretical value are still less, and the fluctuation of the Noise Variance Estimation value between each frame data is less.For example, be under the situation of 18dB in signal to noise ratio, deviation and fluctuation are compared with utilizing the resulting result of blind estimating method, all make moderate progress, and estimated performance is significantly improved.

The Noise Variance Estimation method that various embodiments of the present invention provide, not only can be applied in the aviation wide-band mobile communication technology, can also be applied in traditional moving communicating field, for example, switch and the identification of modulation signal etc. need be carried out under the application scenarios of Noise Variance Estimation.

The Noise Variance Estimation method that the embodiment of the invention provides, utilize the channel frequency response that receives pilot signal in the signal, calculate the peak value of the channel gain of pilot signal in each OFDM symbol, and then according to the peak value of the channel gain of pilot signal in each OFDM symbol, second moment and the Fourth-order moment of whole signals in each OFDM symbol, calculate the estimated value of the noise variance of whole signals in each OFDM symbol, by carrying out in the estimation approach to received signal noise variance, considered the influence of channel condition to the Noise Variance Estimation value, even make under the condition of high s/n ratio, the deviation of the estimated value of middle noise variance and theoretical value also can be smaller to received signal, improved effectively noise variance is carried out estimation performance.

One of ordinary skill in the art will appreciate that: all or part of step that realizes above-mentioned each 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 above-mentioned each 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 each embodiment is not intended to limit only in order to technical scheme of the present invention to be described; Although the present invention has been described in detail with reference to aforementioned each embodiment, 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 puts down in writing, and perhaps some or all of 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 scope of various embodiments of the present invention technical scheme.

Claims (6)

1. a Noise Variance Estimation method is characterized in that, comprising:

Channel frequency response according to pilot signal at least one target sub-carriers in the orthogonal frequency division multiplex OFDM symbol, calculate the amplitude of the channel gain of pilot signal in each target sub-carriers, described target sub-carriers is the subcarrier that is used for the carrying pilot signal in the described OFDM symbol;

According to the amplitude of the channel gain of pilot signal in described at least one target sub-carriers, obtain the peak value of the channel gain of pilot signal in the described OFDM symbol;

According to whole whole Fourth-order moment of signals in the second moment of signals and the described OFDM symbol in the peak value of the channel gain of pilot signal in the described OFDM symbol, the described OFDM symbol, obtain in the described OFDM symbol all estimated values of the noise variance of signals.

2. Noise Variance Estimation method according to claim 1 is characterized in that, described channel frequency response according to pilot signal in the target sub-carriers calculates before the amplitude of the channel gain of pilot signal in the described target sub-carriers, and described method also comprises:

In at least one OFDM symbol from the every frame signal that receives, obtain described at least one target sub-carriers in described at least one OFDM symbol respectively;

Calculate the channel frequency response of pilot signal in each target sub-carriers in described at least one OFDM symbol;

Correspondingly, after the estimated value of the noise variance of whole signals, described method also comprises in the described OFDM symbol of described acquisition:

Utilize the described every frame signal that receives, and whole estimated value of the noise variance of signals at least one the OFDM symbol in every frame signal, restore the signal that transmitter is launched.

3. Noise Variance Estimation method according to claim 1 and 2, it is characterized in that, the channel frequency response of pilot signal is in the described target sub-carriers, the ratio between the amplitude theoretical value of the amplitude of the actual pilot signal that receives on the described target sub-carriers and described pilot signal.

4. Noise Variance Estimation method according to claim 1 is characterized in that, described channel frequency response according to pilot signal in the target sub-carriers, and the amplitude of calculating the channel gain of pilot signal in the described target sub-carriers comprises:

Calculate the mould value of the channel frequency response of pilot signal in the described target sub-carriers;

With the mould value of the channel frequency response of pilot signal in the described target sub-carriers, as the amplitude of the channel gain of pilot signal in the described target sub-carriers.

5. Noise Variance Estimation method according to claim 1 is characterized in that, described amplitude according to the channel gain of pilot signal in described at least one target sub-carriers, and the peak value that obtains the channel gain of pilot signal in the described OFDM symbol is specially:

$k_{h,n}=\frac{E\left\{{\left|H_{n,j}\right|}^4\right\}}{E{\left\{{\left|H_{n,j}\right|}^2\right\}}^2};$

Wherein, | H _{N, j}| be the amplitude of the channel gain of pilot signal in j the target sub-carriers in n the OFDM symbol, E{|H _{N, j}| ⁴To each target sub-carriers in n the OFDM symbol | H _{N, j}| biquadratic carry out average computing, E{|H _{N, j}| ²To each target sub-carriers in n the OFDM symbol | H _{N, j}| quadratic power carry out average computing, k _{H, n}It is the peak value of the channel gain of pilot signal in n the OFDM symbol.

6. Noise Variance Estimation method according to claim 1 is characterized in that, the second moment M of whole signals in the described OFDM symbol _{2, n}For,

The Fourth-order moment M of whole signals in the described OFDM symbol _{4, n}For, $M_{4,n}=E\left\{{\left(y_{n,k}{y}_{n,k}^{*}\right)}^2\right\};$

Wherein, y _{N, k}Be the signal in k the subcarrier in whole subcarriers of n OFDM symbol,

Be that product to the signal in each subcarrier in n the OFDM symbol and conjugated signal carries out the average computing;

Be the quadratic power to the product of the signal in each subcarrier in n the OFDM symbol and conjugated signal, the value of averaging computing;

Correspondingly, according to whole whole Fourth-order moment of signals in the second moment of signals and the described OFDM symbol in the peak value of the channel gain of pilot signal in the described OFDM symbol, the described OFDM symbol, obtaining in the described OFDM symbol all, the estimated value of the noise variance of signals is specially:

Wherein, N _nIt is the estimated value of the noise variance of whole signals in n the OFDM symbol.