CN106302286A - A kind of signal to noise ratio blind estimating method of ofdm system - Google Patents

Abstract

The present invention relates to the signal to noise ratio blind estimating method of a kind of ofdm system, belong to wireless communication technology field.The signal to noise ratio blind estimating method leading symbol based on ofdm system signal frame of this ofdm system obtains data subcarrier mean power and virtual carrier mean power and then determines signal to noise ratio.Therefore, than in prior art, the method for estimation of signal to noise ratio based on pilot aided has higher communication efficiency；Higher than the estimated accuracy of the method using virtual carrier to carry out signal-to-noise ratio (SNR) estimation, meanwhile, the signal to noise ratio blind estimating method complexity of the ofdm system of the present invention is low, it is achieved mode is simple, it is achieved with low cost.

Description

A kind of signal to noise ratio blind estimating method of ofdm system

Technical field

The present invention relates to wireless communication technology field, special ofdm system technical field, specifically refer to a kind of ofdm system Signal to noise ratio blind estimating method.

Background technology

Signal to noise ratio is the key parameter of signal of communication, and it is one of measurement index of communication quality, many fields in wireless communications Close, such as identification, the iterative decoding of Turbo Code of modulated signal, the power in mobile communication controls, Adaptive Modulation switches, Self adaptation handovers etc., are required for knowing the numerical value of signal to noise ratio, to obtain optimal performance.Therefore, in self adaptation OFDM Communication system (OFDM, Orthogonal Frequency Division Multiplexing, orthogonal frequency division multiplexi) in, Signal to noise ratio is the important evidence realizing multinomial communication technology.

The method of ofdm system signal-to-noise ratio (SNR) estimation may be roughly divided into two classes at present: a class is noise based on pilot data auxiliary Rate estimation, a class is blind estimating method.For slowly varying channel, the former needs to send out the most on each subcarrier Sending pilot data, this will reduce communication system information efficiency of transmission；And blind estimating method is built upon useful signal and noise jamming On the basis of statistical property, pilot data need not be sent, thus its communication efficiency is higher, but it estimates that performance is not as the former.For Do not reduce communication system information efficiency of transmission and improve signal-to-noise ratio (SNR) estimation precision, typically using blind estimating method, and seek more preferably Method with improve estimate performance.

Recent year external Wireless OFDM System signal-to-noise ratio (SNR) estimation aspect has carried out some research.Wherein better performances Boumard method and employing virtual carrier method etc..Boumard method utilizes leading symbol to estimate noise variance, utilizes son letter Road estimation coefficient obtains system signal noise ratio further, but performance is affected by delay spread；The performance of virtual carrier method can be had The impact of body system filter.

Summary of the invention

It is an object of the invention to overcome above-mentioned shortcoming of the prior art, it is provided that a kind of based on ofdm system signal frame leading Symbol realizes, and complexity is low, and signal-to-noise ratio (SNR) estimation performance is good, and precision is higher, and implementation is simple, it is achieved with low cost The signal to noise ratio blind estimating method of ofdm system.

In order to realize above-mentioned purpose, the signal to noise ratio blind estimating method employing following steps of the ofdm system of the present invention:

(1) OFDM leading symbol is carried out Fourier transformation；

(2) data subcarrier mean power P is obtained according to the leading symbol after being fourier transformed_s；

(3) virtual carrier mean power P is obtained according to the leading symbol after being fourier transformed_v；

(4) determine that signal to noise ratio is according to described average power signal and described virtual carrier mean power

In the signal to noise ratio blind estimating method of this ofdm system, described step (4) particularly as follows:

With the average signal-to-noise ratio determined according to following formulaSignal to noise ratio as described ofdm system:

In above formula,For the channel on kth subcarrier, determine according to following formula:

In above formula, M_{2, k}For reception data Y on the kth subcarrier of m-th OFDM leading symbol_{M, k}Second moment describe, W is noise variance.

In the signal to noise ratio blind estimating method of this ofdm system, the kth subcarrier of described m-th OFDM leading symbol On reception data Y_{M, k}Second moment describeDetermine according to following formula:

In above formula, L is the number of OFDM leading symbol, | | | | represent delivery.

In the signal to noise ratio blind estimating method of this ofdm system, described noise variance W determines according to following formula:

$W=\frac{1}{2}E\left\{{\left|\right|Y_{m,k}^{\′}-Y_{m+1,k}^{\′}\left|\right|}^2\right\}$

In above formula, | | | | represent delivery.E{. | k} represents and seeks conditional expectation, m=0,1 ..., L-1, L are OFDM leading symbol Number.

In the signal to noise ratio blind estimating method of this ofdm system, described leading symbol includes short training symbols and long training symbol.

In the signal to noise ratio blind estimating method of this ofdm system,

Described step (1) particularly as follows:

(1-1) short training symbols is divided into two groups, carries out Fourier transformation respectively；

Described step (2) particularly as follows:

(2-1) short training symbols after being fourier transformed according to one of which obtains data subcarrier mean power P_s；

Described step (3) particularly as follows:

(3-1) virtual carrier mean power P is obtained according to the short training symbols after wherein another group is fourier transformed_v。

In the signal to noise ratio blind estimating method of this ofdm system,

Described step (1) particularly as follows:

(1-2) two long training symbols are carried out Fourier transformation respectively；

Described step (2) particularly as follows:

(2-2) data subcarrier mean power P is obtained according to the long training symbol after being fourier transformed_s；

Described step (3) particularly as follows:

(3-2) the long training symbol after being fourier transformed two subtracts each other, and the power of the signal obtained is noise power in channel Twice, and then noise power can be obtained.

In the signal to noise ratio blind estimating method of this ofdm system,

Described step (1) particularly as follows:

(1-3) short training symbols is divided into two groups, carries out Fourier transformation respectively, and two long training symbols are carried out in Fu respectively Leaf transformation；

Described step (2) particularly as follows:

(2-3) short training symbols after being fourier transformed according to one of which obtains data subcarrier mean power P_s；

Described step (3) particularly as follows:

(3-3) the long training symbol after being fourier transformed two subtracts each other, and the power of the signal obtained is noise power in channel Twice, and then noise power can be obtained.

In the signal to noise ratio blind estimating method of this ofdm system,

Described step (1) particularly as follows:

(1-4) short training symbols is divided into two groups, carries out Fourier transformation respectively, and two long training symbols are carried out in Fu respectively Leaf transformation；

Described step (2) particularly as follows:

(2-4) data subcarrier mean power P is obtained according to the long training symbol after being fourier transformed_s；

Described step (3) particularly as follows:

(3-4) short training symbols after being fourier transformed according to a group obtains virtual carrier mean power P_v。

Have employed the signal to noise ratio blind estimating method of the ofdm system of this invention, due to it based on ofdm system signal frame before Lead symbol obtain data subcarrier mean power and virtual carrier mean power and then determine signal to noise ratio, therefore, than base in prior art Method of estimation in the signal to noise ratio of pilot aided has higher communication efficiency；Than the method using virtual carrier to carry out signal-to-noise ratio (SNR) estimation Estimated accuracy higher, meanwhile, the signal to noise ratio blind estimating method complexity of the ofdm system of the present invention is low, it is achieved mode is simple, Realize with low cost.

Accompanying drawing explanation

Fig. 1 is the step schematic diagram of the signal to noise ratio blind estimating method of the ofdm system of the present invention.

Fig. 2 is the OFDM leading symbol structural representation of the ofdm system that the present invention uses.

Fig. 3 is the mean square error curve in the case of four kinds of signal-noise ratio estimation methods of the present invention.

In figure, a-only uses short leading estimation；B-only uses long preambles to estimate；C-uses short leading and long preambles combination estimation； D-uses short leading (virtual carrier) and long preambles to estimate.

Fig. 4 is output signal-to-noise ratio estimation curve in the case of setting input signal-to-noise ratio.

Detailed description of the invention

In order to be more clearly understood that the technology contents of the present invention, describe in detail especially exemplified by following example.

Refer to shown in Fig. 1, for the step schematic diagram of signal to noise ratio blind estimating method of the ofdm system of the present invention.

In one embodiment, the signal to noise ratio blind estimating method of ofdm system, it is characterised in that described method include with Lower step:

(1) OFDM leading symbol is carried out Fourier transformation；

(2) data subcarrier mean power P is obtained according to the leading symbol after being fourier transformed_s；

(3) virtual carrier mean power P is obtained according to the leading symbol after being fourier transformed_v；

(4) determine that signal to noise ratio is according to described average power signal and described virtual carrier mean power

In one more preferably embodiment, described step (4) particularly as follows:

With the average signal-to-noise ratio determined according to following formulaSignal to noise ratio as described ofdm system:

In above formula,For the channel on kth subcarrier, determine according to following formula:

In above formula, M_{2, k}For reception data Y on the kth subcarrier of m-th OFDM leading symbol_{M, k}Second moment describe, W is noise variance.

In a kind of further preferred embodiment, connecing on the kth subcarrier of described m-th OFDM leading symbol Receive data Y_{M, k}Second moment describeDetermine according to following formula:

In above formula, L is the number of OFDM leading symbol, | | | | represent delivery.

Described noise variance W determines according to following formula:

$W=\frac{1}{2}E\left\{{\left|\right|Y_{m,k}^{\′}-Y_{m+1,k}^{\′}\left|\right|}^2\right\}$

In above formula, | | | | represent delivery.E{. | k} represents and seeks conditional expectation, m=0,1 ..., L-1, L are OFDM leading symbol Number.

In another kind of further preferred embodiment, described leading symbol includes short training symbols and long training symbol.Before Lead that the selection of symbol can use in following four preferred implementation any one.

The first preferred embodiment, described step (1)～(3) particularly as follows:

(1-1) short training symbols is divided into two groups, carries out Fourier transformation respectively；

(2-1) short training symbols after being fourier transformed according to one of which obtains data subcarrier mean power P_s；

(3-1) virtual carrier mean power P is obtained according to the short training symbols after wherein another group is fourier transformed_v。

The second preferred embodiment, described step (1)～(3) particularly as follows:

(1-2) two long training symbols are carried out Fourier transformation respectively；

(2-2) data subcarrier mean power P is obtained according to the long training symbol after being fourier transformed_s；

(3-2) the long training symbol after being fourier transformed two subtracts each other, and the power of the signal obtained is noise power in channel Twice, and then noise power can be obtained.

The third preferred embodiment, described step (1)～(3) particularly as follows:

(1-3) short training symbols is divided into two groups, carries out Fourier transformation respectively, and two long training symbols are carried out in Fu respectively Leaf transformation；

(2-3) short training symbols after being fourier transformed according to one of which obtains data subcarrier mean power P_s；

(3-3) the long training symbol after being fourier transformed two subtracts each other, and the power of the signal obtained is noise power in channel Twice, and then noise power can be obtained.

4th kind is preferred embodiment, described step (1)～(3) particularly as follows:

(1-4) short training symbols is divided into two groups, carries out Fourier transformation respectively, and two long training symbols are carried out in Fu respectively Leaf transformation；

(2-4) data subcarrier mean power P is obtained according to the long training symbol after being fourier transformed_s；

(3-4) short training symbols after being fourier transformed according to a group obtains virtual carrier mean power P_v。

In actual applications, the signal to noise ratio blind estimating method of the ofdm system of the method for the present invention specifically comprises the following steps that

The present invention uses the effective durations of OFDM symbol to be T_U=80 μ s and total number of sub-carriers are the ofdm system of 1024. Short leading and long preambles all uses CAZAC sequence, and its Cyclic Prefix (CP) durations is i.e. T_CP=20 μ s, long preambles Symbol cyclic prefix duration T_CP1=40us, short leading duration T_short=10*T_CP=200us, useful subcarrier Number N_S=480, FFT/IFFT transform length is 1024, carries a width of 6MHz, as shown in Figure 2.

According to the white Gaussian noise on any subcarrier, there is identical statistical property (it is unrelated with subcarrier sequence number) and leading The feature that in OFDM symbol, data modes of emplacement is identical with OFDM data symbol.First the leading symbol received is entered Row FFT, calculates data subcarrier and the mean power of virtual carrier the most respectively, i.e. can get signal-to-noise ratio (SNR) estimation value:

$S\hat{N}R=\frac{P_s}{P_v}---\left(1\right)$

In formula, P_sFor the mean power on data symbol subcarrier on leading symbol, and P_vFor on virtual carrier on targeting sequencing symbol Average power signal, think at this for noise average power.Process above process can represent with Fig. 1.

In an ofdm system, signal frame structure often uses preposition leading symbol to add the structure of data, as at WLAN In IEEE802.11a standard, leading symbol includes the short training symbols of 10 cycles repetitions and the long training symbol of 2 cycle repetitions Number, mark space is identical with normal OFDM symbol.As shown in Figure 2.So proposing four kinds of signal-to-noise ratio (SNR) estimation schemes:

1) short preamble symbols is only used

Short training symbols is by CAZAC sequence modulation to carrier wave, then short training symbols is divided into two groups, is FFT respectively Conversion, statistical data carrier wave and the mean power of virtual carrier, can calculate signal to noise ratio the most respectively.

2) long preambles symbol is only used

Two long preambles symbols are by CAZAC sequence modulation to carrier wave, do FFT the most respectively, owing to long preambles accords with Not having virtual carrier in number, subtracted each other by the long preambles symbol after FFT, the power of the signal obtained is noise power in channel Twice, and then noise power can be obtained, thus calculate signal to noise ratio.

3) short preamble symbols and long preambles symbol combine

Use both the above method simultaneously, calculate data carrier power and noise power, thus calculate signal to noise ratio.

4) short preamble symbols (only using virtual carrier) and long preambles symbol combine

Similar with the third method, short preamble symbols only uses virtual carrier, does not use data carrier, because in short preamble symbols Data carrier negligible amounts, calculates time error bigger.

In an ofdm system, in frequency domain, m-th transmission symbol can be expressed as:

C_m=[c_{M, 0}, c_{M, 1}..., c_{M, k}..., c_{M, N-1}] (2)

In above formula, N represents IFFT length, c_{M, k}Represent the modulated data on m-th OFDM symbol kth subcarrier.? Adjusting data is after channel, at receiving terminal after FFT processes, and the reception number on m-th OFDM symbol kth subcarrier According to being given below:

Y_{M, k}=c_{M, k}H_{M, k}+n_{M, k} (3)

N in above formula_{M, k}Being zero-mean, variance is separate between the white Gaussian noise of W, and each subcarrier.H_{M, k}For kth The channel coefficients of individual subcarrier, can be expressed as:

$H_{m,k}=\underset{l=1}{\overset{G}{\mathrm{\Σ}}}h(\mathrm{mT},l{T}_s)\·e^{-j2\mathrm{\π}\frac{\mathrm{kl}}{N}}---\left(4\right)$

H (mT, lT_s) be m-th OFDM wireless channel impulse response discrete sampling, T is the OFDM symbol cycle, T_s Being the sampling period of receiving terminal, G is the sampling length of wireless channel impulse response.

In Wireless OFDM System, the persistent period of OFDM bag is shorter than channel coherency time, then H_{M, k}Can be by H_kReplace. Then for the l-th symbol of OFDM, the SNR ρ of kth subcarrier_kIt is given below:

${\mathrm{\ρ}}_k=\frac{E\left\{{\left|\right|c_{m,k}\left|\right|}^2\right|k\}{\left|\right|H_k\left|\right|}^2}{W}---\left(5\right)$

In above formula, | | | | represent delivery.E{. | k} represents and seeks conditional expectation.M=0,1 ..., L-1.

Then average SNR is given below:

${\mathrm{\ρ}}_{\mathrm{av}}=\frac{E\left\{E\right[{\left|\right|c_{m,k}\left|\right|}^2\left|k\right]{\left|\right|H_k\left|\right|}^2\}}{W}=E\left\{{\mathrm{\ρ}}_k\right\}---\left(6\right)$

In above formula, k=0,1 ..., L-1, m=0,1 ..., L-1.

It is given below: according to formula (3), m-th and the leading of m+1 OFDM training symbol

Y_{M, k}=c_{M, k}H_{M, k}+n_{M, k} (7)

Y_{M+1, k}=c_{M+1, k}H_{M+1, k}+n_{M+1, k} (8)

Here c_{M, k}For given data.If c_{M, k}≠ 0, then suppose | | c_{M, k}| |=1.

Work as c_{M, k}≠ 0, order

$Y_{m,k}^{\′}=Y_{m,k}\·c_{m,k}^{*}=H_k+n_{m,k}\·c_{m+1,k}^{*}=H_k+n_{m,k}^{\′}---\left(9\right)$

$Y_{m+1,k}^{\′}=Y_{m+1,k}\·c_{m,k}^{*}=H_k+n_{m+1,k}\·c_{m+1,k}^{*}=H_k+n_{m+1,k}^{\′}---\left(10\right)$

Work as c_{M, k}=0, order

Y′_{M, k}=Y_{M, k}=n_{M, k}=n '_{M, k} (11)

Y′_{M+1, k}=Y_{M+1, k}=n_{M+1, k}=n '_{M+1, k} (12)

Consider following equation

E{||Y′_{M, k}-Y′_{M+1, k}||²}=E{ | | n '_{M, k}-n′_{M+1, k}||²}=2W (13)

So, noise variance can be estimated as the following formula

$W=\frac{1}{2}E\left\{{\left|\right|Y_{m,k}^{\′}-Y_{m+1,k}^{\′}\left|\right|}^2\right\}---\left(14\right)$

In reality, because the ergodic above formula of noise can be written as

The second moment of equation (3) can be described as

M_{2, k}=E{ | | Y '_{M, k}||²}=E{ | | c_{M, k}H_k||²}+E{||n′_{M, k}||²}=P_k+W (16)

In above formula, P_kIt it is the signal power of kth subcarrier.

From equation (14) and (15), the signal power of kth subcarrier can be described as

P_k=M_{2, k}-W (17)

Then the channel on kth subcarrier is estimated as

Here

In above formula, L is OFDM symbol number.

Then average SNR can be expressed as

In sum, from simulation result Fig. 3 and Fig. 4 it can be seen that the inventive method is feasible, consistent with notional result. The key point of the present invention is:

1) present invention uses OFDM signal to noise ratio blind estimating method based on leading symbol, relative to the method inserting pilot tone, tool Need not overhead, communication efficiency high；

2) present invention uses OFDM signal to noise ratio blind estimating method based on leading symbol, before utilizing relative to Boumard method Leading sign estimation noise variance, utilize subchannel estimation coefficient to obtain system signal noise ratio further, its performance is by the shadow of delay spread Ringing, this method uses the correlation function estimating system signal to noise ratio of leading symbol, has the advantage such as estimation performance and precision height；

3) present invention uses OFDM signal to noise ratio blind estimating method based on leading symbol, relative to using virtual carrier method of estimation, Being affected by concrete system filter, this method uses the method that virtual carrier and data subcarrier combine, and has estimated accuracy high, Realize simple, the features such as complexity is low；

4) present invention uses OFDM signal to noise ratio blind estimating method based on leading symbol, accords with according to short preamble symbols and long preambles Number feature, it is proposed that four kinds of signal-noise ratio estimation methods can be selected according to concrete application scenario, flexibly to reach best estimate Performance and precision.

Have employed the signal to noise ratio blind estimating method of the ofdm system of this invention, due to it based on ofdm system signal frame before Lead symbol obtain data subcarrier mean power and virtual carrier mean power and then determine signal to noise ratio, therefore, than base in prior art Method of estimation in the signal to noise ratio of pilot aided has higher communication efficiency；Than the method using virtual carrier to carry out signal-to-noise ratio (SNR) estimation Estimated accuracy higher, meanwhile, the signal to noise ratio blind estimating method complexity of the ofdm system of the present invention is low, it is achieved mode is simple, Realize with low cost.

In this description, the present invention is described with reference to its specific embodiment.But it is clear that still may be made that various Amendment and conversion are without departing from the spirit and scope of the present invention.Therefore, specification and drawings is considered as illustrative rather than limits Property processed.

Claims (9)

1. the signal to noise ratio blind estimating method of an ofdm system, it is characterised in that described method comprises the following steps:

(1) OFDM leading symbol is carried out Fourier transformation；

(2) data subcarrier mean power P is obtained according to the leading symbol after being fourier transformed_s；

(3) virtual carrier mean power P is obtained according to the leading symbol after being fourier transformed_v；

(4) determine that signal to noise ratio is according to described average power signal and described virtual carrier mean power

The signal to noise ratio blind estimating method of ofdm system the most according to claim 1, it is characterised in that described step (4) particularly as follows:

With the average signal-to-noise ratio determined according to following formulaSignal to noise ratio as described ofdm system:

In above formula,For the channel on kth subcarrier, determine according to following formula:

In above formula, M_2,kFor reception data Y on the kth subcarrier of m-th OFDM leading symbol_m,kSecond moment describe, W is noise variance.

The signal to noise ratio blind estimating method of ofdm system the most according to claim 2, it is characterised in that described m Reception data Y on the kth subcarrier of individual OFDM leading symbol_m,kSecond moment describeDetermine according to following formula:

In above formula, L is the number of OFDM leading symbol, | | | | represent delivery.

The signal to noise ratio blind estimating method of ofdm system the most according to claim 2, it is characterised in that described noise Variance W determines according to following formula:

$W=\frac{1}{2}E\left\{\right||Y_{m,k}^{\′}-Y_{m+1,k}^{\′}{\left|\right|}^2\}$

In above formula, | | | | represent delivery.E{. | k} represents and seeks conditional expectation, m=0,1 ..., L-1, L are OFDM leading symbol Number.

The signal to noise ratio blind estimating method of ofdm system the most according to any one of claim 1 to 4, it is characterised in that Described leading symbol includes short training symbols and long training symbol.

The signal to noise ratio blind estimating method of ofdm system the most according to claim 5, it is characterised in that

Described step (1) particularly as follows:

(1-1) short training symbols is divided into two groups, carries out Fourier transformation respectively；

Described step (2) particularly as follows:

(2-1) short training symbols after being fourier transformed according to one of which obtains data subcarrier mean power P_s；

Described step (3) particularly as follows:

(3-1) virtual carrier mean power P is obtained according to the short training symbols after wherein another group is fourier transformed_v。

The signal to noise ratio blind estimating method of ofdm system the most according to claim 5, it is characterised in that

Described step (1) particularly as follows:

(1-2) two long training symbols are carried out Fourier transformation respectively；

Described step (2) particularly as follows:

(2-2) data subcarrier mean power P is obtained according to the long training symbol after being fourier transformed_s；

Described step (3) particularly as follows:

(3-2) the long training symbol after being fourier transformed two subtracts each other, and the power of the signal obtained is noise power in channel Twice, and then noise power can be obtained.

The signal to noise ratio blind estimating method of ofdm system the most according to claim 5, it is characterised in that

Described step (1) particularly as follows:

(1-3) short training symbols is divided into two groups, carries out Fourier transformation respectively, and two long training symbols are carried out in Fu respectively Leaf transformation；

Described step (2) particularly as follows:

(2-3) short training symbols after being fourier transformed according to one of which obtains data subcarrier mean power P_s；

Described step (3) particularly as follows:

(3-3) the long training symbol after being fourier transformed two subtracts each other, and the power of the signal obtained is noise power in channel Twice, and then noise power can be obtained.

The signal to noise ratio blind estimating method of ofdm system the most according to claim 5, it is characterised in that

Described step (1) particularly as follows:

(1-4) short training symbols is divided into two groups, carries out Fourier transformation respectively, and two long training symbols are carried out in Fu respectively Leaf transformation；

Described step (2) particularly as follows:

(2-4) data subcarrier mean power P is obtained according to the long training symbol after being fourier transformed_s；

Described step (3) particularly as follows:

(3-4) short training symbols after being fourier transformed according to a group obtains virtual carrier mean power P_v。