CN101616105B - Method for estimating channel of orthogonal frequency division multiplexing system - Google Patents

Abstract

The invention relates to a method for estimating a channel of an orthogonal frequency division multiplexing system, comprising the following steps: dividing received continuous data into a plurality of time-frequency two-dimensional data blocks; utilizing pilot frequencies to obtain channel estimation values of pilot frequency positions and channel estimation values of all carrier wave frequency positions of the time-frequency two-dimensional data blocks; obtaining a time-domain corrected value according to a difference value of the channel estimation values of the pilot frequency positions and the channel estimation values of the carrier wave frequency positions of a carrier wave locating the pilot frequency; and obtaining channel estimation values of all data of all the time-frequency two-dimensional data blocks by adding the channel estimation values of all the carrier wave frequency positions and the time-domain corrected value. The method sufficiently utilizes the frequency correlation and the time correlation of a wireless channel, thereby more accurately estimating the shock response of the channel and also reducing the complexity with little influence due to the small number of the pilot frequencies.

Description

The method of channel estimating in a kind of ofdm system

Technical field

The present invention relates to mobile radio system, be specifically related to the method for channel estimating in a kind of ofdm system.

Background technology

Along with through the video that the Internet carried out, the phenomenal growth of voice communications versus data communications flow, and the popularizing fast of mobile phone, the user hopes that more urgently GSM can provide the access of mobile multi-media service.Therefore countries in the world are also all invested multiple access access mobile communication system to sight, and wherein ofdm/ofdma (OFDM/OFDMA) is exactly one of a kind of noticeable candidate scheme.

At the receiving terminal of ofdm system, in order to improve the performance of system, adopt coherent demodulation usually, coherent demodulation needs channel parameter information, and it can obtain through channel estimating.Performance for estimating channel directly influences systematic function, is one of key technology of receiver.Channel estimation methods based on pilot tone has obtained wide influence.Under multipath fading and Quick-Change channel situation, for example in the High Data Rate OFDM mobile system,, generally adopt comb pilot signal to estimate to obtain the Real-time Channel response for overcoming the adverse effect of channel.

The Comb Pilot channel estimating is made up of pilot sub-carrier channel estimation method and difference algorithm.Because the channel response value at data subcarrier is to obtain through the channel response value of pilot frequency carrier wave is carried out interpolation, so systematic function depends on the accuracy that the pilot sub-carrier channel response is estimated to a great extent.Channel estimating based on pilot tone has fairly simple LS (Least Square; Least square) estimates (to select from " statistics of signal detects and estimation theory "; Li Daoben is write); More complicated MMSE (Minimum Mean Square Error, least mean-square error) estimates methods such as (select from " statistics of signal detects and estimation theory ", Li Daoben are write).

Traditional least square LS algorithm mainly contains the match of (one) frequency and two kinds of methods of (two) time domain and frequency domain combined match.

(1) the frequency domain fitting algorithm mainly is the channel estimating of carrying out pilot frequency carrier wave according to known pilot, utilizes the channel estimation value of pilot frequency carrier wave to carry out the channel response value that the linear match of frequency domain one dimension obtains the data carrier position.Key step comprises:

1) data with the continuous N that a receives symbol are divided into plurality of time-frequency two-dimensional data blocks, and each data block takies m symbol in time domain, take n carrier wave on the frequency domain, so each data block comprises m * n data;

2) utilize pilot tone in each data block to obtain the channel estimation value

of pilot sub-carrier position

3) For each data block locations pilot carrier channel response value

frequency-domain fitting algorithm to obtain the entire data block locations all carriers channel response value

Its frequency domain fitting algorithm mainly comprises:

H (k) is the channel frequency response at subcarrier k place, and P is the location index collection of pilot sub-carrier in the data block.H (k) is by estimating like drag:

$\hat{H}\left(k\right)=c_{1}k+c_2=C^H{q}_k=q_k^T{C}^{*}$

Wherein: $q_k^T=[k,1]$

C ^H＝[c ₁，c ₂]

The principle of confirming C is: $\underset{c}{\mathrm{Min}}\left\{\underset{k\∈P}{\mathrm{\Σ}}{|H\left(k\right)-C^H{q}_k|}^2\right\}$

4) obtain the channel response value of the data carrier position of each symbol according to the channel response value of all carrier positions of whole data block;

That is, suppose D _iBe the location index of the data subcarrier of i symbol in the data block, then the channel response value of the data subcarrier of i symbol does

Can find out that by above derivation the temporal correlation of channel is not considered in the frequency match, think that the channel response value of all symbol same carrier wave positions is all identical in each data block; So good performance is arranged under the channel condition of low-speed motion, this is because under the channel of low-speed motion, the coefficient correlation of interchannel is bigger; Estimated channel and actual channel error are very little, owing to calculate simply, are well suited for being applied in the actual channel; But under the channel of high-speed motion; The temporal correlation of interchannel is relatively poor, and the temporal correlation of channel is very big to the influence of channel estimating, so traditional LS frequency fitting algorithm performance is poor under high-speed motion.

(2) time domain and frequency domain combined fitting algorithm mainly is the channel estimating of carrying out pilot frequency carrier wave according to known pilot, utilizes the channel estimation value of pilot frequency carrier wave to carry out the channel response value that time domain and the linear match of frequency domain two dimension obtain each symbol data carrier position.Key step comprises:

The data of the continuous N that a receives symbol are divided into plurality of time-frequency two-dimensional data blocks, and each data block takies m symbol in time domain, takies n carrier wave on the frequency domain, so each data block comprises m * n data;

Utilize pilot tone in each data block to obtain the channel estimation value

of pilot sub-carrier position

For each data block locations pilot carrier channel response value frequency-domain fitting algorithm entire block of data symbols the data carrier positions of all the channel response values

Its frequency domain fitting algorithm mainly comprises:

(k n) is the channel frequency response at n symbol subcarrier k place to H, and P is the location index collection of pilot sub-carrier in the data block.

H (k, n) by estimating like drag: $\hat{H}(k,n)=c_{1}k+c_{2}n+c_3=C^H{q}_k=q_k^T{C}^{*}$

Wherein: $q_{(k,n)}^T=[k,n,1]$

C ^H＝[c ₁，c ₂，c ₃]

The principle of confirming C is: $\underset{c}{\mathrm{Min}}\left\{\underset{k\∈P}{\mathrm{\Σ}}{|H(k,n)-C^H{q}_{k,n}|}^2\right\}$

Find out by above; Though time domain and frequency domain combined fitting algorithm has been considered the temporal correlation and the frequency dependence of channel simultaneously, because the time-frequency domain fitting algorithm utilizes plane of limited pilot tone match to obtain the channel response value of data carrier, so the pilot number of the performance of match and each symbol has much relations; Pilot number is abundant; The performance of match is all right, but greatly reduces band efficiency, and pilot number is few; The performance of match even than the poor performance of frequency match, so time domain and frequency domain combined match generally is applied in the more allocation of carriers mode of pilot number.

In addition; The MMSE algorithm need be known correlation between channels in advance, and the relevant information of channel belongs to the second-order statistics of channel status, and statistical property is all handled through time average in reality; This temporal on average replace statistical on average only when channel belongs to stationary random process, be only no inclined to one side; And channel generally is not a stationary random process, and error has been introduced in this processing, and the MMSE algorithm also must estimate signal to noise ratio in addition; Whole estimation procedure complexity is very high, is not suitable for real system.

Summary of the invention

The technical issues that need to address of the present invention are the methods how channel estimating in a kind of ofdm system is provided, and under the prerequisite of higher band efficiency, can obtain the channel response of whole channel with lower complexity more accurately.

Above-mentioned technical problem of the present invention solves like this, and the method for channel estimating in a kind of ofdm system is provided, and may further comprise the steps:

1.1) continuous data that receives is divided into some time-frequency two-dimensional data blocks, that is: each time-frequency two-dimensional data blocks takies identical duration frequency range;

1.2) utilize pilot tone to obtain the channel estimation value that channel estimation values of pilot frequency positions also further obtains all carrier frequency positions of said two-dimensional blocks of data;

1.3) obtain the time domain correction value according to the difference of the channel estimation value of the carrier frequency position of channel estimation values of pilot frequency positions and this pilot tone place carrier wave;

1.4) according to said step 1.2) and in the channel estimation value of all carrier frequency positions add the above step 1.3) in the time domain correction value obtain the channel estimation value of all data in the said data block.

According to the method for channel estimating provided by the invention, specific as follows:

2.1) data of the continuous N that a receives symbol are divided into plurality of time-frequency two-dimensional data blocks, each data block takies m symbol in time domain, takies n carrier wave on the frequency domain, comprises m * n data, and wherein: m, n are natural number;

2.2) utilize the pilot tone in each data block to obtain channel estimation values of pilot frequency positions

2.3) to the location of the pilot channel estimation value

to obtain the data block based on the carrier frequency positions of all the channel estimation value

2.4) according to channel estimation values of pilot frequency positions Channel estimation value with the carrier frequency position of this pilot tone place carrier wave Difference obtain each symbol time domain correction value w _i

2.5) according to the channel estimation value of said all carrier frequency positions

Add each symbol time domain correction value w _iObtain the channel estimation value of all data in the said data block

According to the method for channel estimating provided by the invention, said step 2.2) carries out the channel estimating of said pilot frequency locations with the least-squares estimation method.

According to the method for channel estimating provided by the invention, said channel estimation values of pilot frequency positions

According to formula

Wherein: P _TransBe the pilot frequency sequence that sends, P _ReceiveBe the pilot frequency sequence that receives, p is the set of the corresponding frequency domain sequence number of pilot tone in the data block, and k ' is the arrangement sequence number of pilot tone in the pilot set.

According to the method for channel estimating provided by the invention, said step 2.3) carries out the channel estimating of said all carrier frequency positions with the least-squares estimation method.

Method according to channel estimating provided by the invention; The channel estimation value of said carrier frequency position

is that basis is obtained with drag, and k is the corresponding frequency domain sequence number of each symbol of data block:

$\hat{H}\left(k\right)=c_{1}k+c_2=C^H{q}_k=q_k^T{C}^{*}$

Wherein: $q_k^T=[k,1]$

C ^H＝[c ₁，c ₂]

The principle of confirming C is: $\underset{c}{\mathrm{Min}}\left\{\underset{k\∈P}{\mathrm{\Σ}}{|H\left(k\right)-C^H{q}_k|}^2\right\}.$

According to the method for channel estimating provided by the invention, the corresponding symbol time domain of each time slot correction value w in the said data block _iGet said step 1.4) in the average of difference.

Method according to channel estimating provided by the invention; Said step 1.4) channel estimation value of all data in

obtains according to formula ; Wherein: i is the corresponding time domain sequence numbers of each data of data block, and k is the corresponding frequency domain sequence numbers of each data of data block.

Method according to channel estimating provided by the invention; This method comprises that also the channel response value

of utilizing all data carries out channel equalization to receiving data block, and then balanced data is carried out coherent demodulation.

According to the method for channel estimating provided by the invention, said ofdm system is based on the ofdm system of IEEE802.16a/802.16d/802.16e/LTE standard.

The method of channel estimating in a kind of ofdm system provided by the invention is in the channel estimating of the carrier position that obtains

The basis on again according to the channel estimating of pilot frequency locations

Obtain time domain correction value w _iFurther obtain the channel response of whole channel, made full use of the frequency dependence and the temporal correlation of wireless channel like this, estimate the impulse response of channel more accurately

And complexity is low, and also influence is little less for the pilot tone number.

Description of drawings

Further the present invention is elaborated below in conjunction with accompanying drawing and specific embodiment.

Fig. 1 is the structural representation of the data block of first embodiment of the invention use;

The flow chart that Fig. 2 realizes for the present invention;

Fig. 3 is the comparison diagram of the channel estimating performance of first embodiment of the invention and traditional LS algorithm;

Fig. 4 is the structural representation of the data block of second embodiment of the invention use.

Embodiment

At first, brief description key step of the present invention specifically comprises:

1. the data with the continuous N that a receives symbol are divided into plurality of time-frequency two-dimensional data blocks, and each data block takies m symbol in time domain, take n carrier wave on the frequency domain, so each data block comprises m * n data;

2. utilize the pilot tone in each data block to obtain channel estimation values of pilot frequency positions;

3. utilize the channel response value of pilot frequency locations in each data block to obtain the channel response value of all carrier positions in the data block;

4. with the channel response value of the carrier position in each data block deduct step 3. in the channel response value of pilot frequency locations on the same frequency, and ask for the average of difference, obtain the time domain correction value of this symbol response;

5. the channel response value that step is obtained carrier position in 3. adds the time domain correction value of each the symbol response that obtains during step 4., obtains the channel response value of all data in each data block;

In addition, comprise that also the channel response value of utilizing each symbol carries out channel equalization to the data that receive, and then balanced data is carried out coherent demodulation;

The channel estimation methods of described ofdm system, wherein, said step 1. in; The criteria for classifying of data block depends on the allocation of carriers mode that real system uses, and for the up continuous sub-carrier distribution manner of 802.16e, takies 3 symbols on each data block time domain; Take 18 carrier waves on the frequency domain; And, take 3 symbols on each data block time domain for up part sub-carrier distribution manner, take 4 carrier waves on the frequency domain.

The channel estimation methods of described ofdm system, wherein, said step 2. in, carry out the channel estimating of said pilot sub-carrier position with the least-squares estimation method.

The channel estimation methods of described ofdm system, wherein, said step 3. in, carry out the channel estimating of all sub-carrier positions of said each data block with the least-squares estimation method.

The channel estimation methods of described ofdm system, wherein, said ofdm system is the ofdm system based on the IEEE802.16a/802.16d/802.16e/LTE standard.

Below, be further elaborated in conjunction with accompanying drawing and concrete application of the present invention:

As shown in Figure 1, the data block that first embodiment of the invention is used takies 3 symbol times on time domain, on frequency domain, take 18 carrier spacings.2 pilot sub-carriers and 16 data subcarriers are arranged in each symbol of each data block.The 2nd of first symbol and the 11st carrier position are pilot sub-carrier, and are labeled as P _Recieve(1) and P _Recieve(2); The 5th of second symbol and the 14th carrier position are pilot sub-carrier, and are labeled as P _Recieve(3) and P _Recieve(4); The 8th of the 3rd symbol and the 17th carrier position are pilot sub-carrier, and are labeled as P _Recieve(5) and P _Recieve(6).See that from frequency pilot sub-carrier takies bandwidth seldom, so bandwidth availability ratio is very high.

As shown in Figure 2, be the flow chart of channel estimation methods of the present invention, the supposing the system ideal synchronisation, the concrete steps of this embodiment channel estimating are following:

202) data that receive (comprising pilot tone and data) are divided into data block, each data block takies 3 symbol times on time domain, on frequency domain, takies 18 carrier spacings.The method of salary distribution of pilot sub-carrier and data subcarrier is as shown in Figure 1;

204) pilot sub-carrier of each data block is utilized the channel response value of LS algorithm estimating pilot frequency sub-carrier positions, promptly

${\hat{H}}_p=\frac{P_{\mathrm{receive}}}{P_{\mathrm{trans}}}=(\frac{P_{\mathrm{receive}}\left(1\right)}{P_{\mathrm{trans}}\left(1\right)},\frac{P_{\mathrm{receive}}\left(2\right)}{P_{\mathrm{trans}}\left(2\right)},\frac{P_{\mathrm{receive}}\left(3\right)}{P_{\mathrm{trans}}\left(3\right)},\frac{P_{\mathrm{receive}}\left(4\right)}{P_{\mathrm{trans}}\left(4\right)},\frac{P_{\text{receive}}\left(5\right)}{P_{\mathrm{trans}}\left(5\right)},\frac{P_{\mathrm{receive}}\left(6\right)}{P_{\mathrm{trans}}\left(6\right)})$

Wherein, P _TransThe pilot frequency sequence that be to send is base station and travelling carriage known sequences all.

206) utilize the LS algorithm to obtain the channel response value of all carrier positions of whole data block to the channel response value

of the pilot sub-carrier position estimating to obtain.

H (k) is the channel frequency response at subcarrier k place, and 1≤k≤18 make p=[2,5,8,11,14,17], and P is the location index collection of pilot sub-carrier in the data block.H (k) is by estimating like drag:

$\hat{H}\left(k\right)=c_{1}k+c_2=C^H{q}_k=q_k^T{C}^{*}$

Wherein: $q_k^T=[k,1]$

C ^H＝[c ₁，c ₂]

The principle of confirming C is: $\underset{c}{\mathrm{Min}}\left\{\underset{k\∈P}{\mathrm{\Σ}}{|H\left(k\right)-C^H{q}_k|}^2\right\}$

$\underset{c}{\mathrm{Min}}\{{|{\hat{H}}_p\left(1\right)-C^H{q}_2|}^2+{|{\hat{H}}_p\left(2\right)-C^H{q}_{11}|}^2+{|{\hat{H}}_p\left(3\right)-C^H{q}_5|}^2+{|{\hat{H}}_p\left(4\right)-{\mathrm{<figure-callout\; id="14"\; label="C\; H\; q"\; filenames="S2008101261573E00031.png"\; state="\{\{state\}\}">C</figure-callout>}}^H{q}_{}{}_{14}|}^2$

$+{|{\hat{H}}_p\left(5\right)-C^H{q}_8|}^2+{|{\hat{H}}_p\left(6\right)-C^H{q}_{17}|}^2\}$

Promptly

Finally calculate:

$\hat{H}\left(k\right)=k\left\{\frac{1}{105}\right[-5{\hat{H}}_p\left(1\right)-3{\hat{H}}_p\left(3\right)-{\hat{H}}_p\left(5\right)+{\hat{H}}_p\left(2\right)+3{\hat{H}}_p\left(4\right)+5{\hat{H}}_p\left(6\right)\left]\right\}$

$+\left\{\frac{1}{105}\right[20{\hat{H}}_p\left(1\right)+19{\hat{H}}_p\left(3\right)+18{\hat{H}}_p\left(5\right)+17{\hat{H}}_p\left(2\right)+16{\hat{H}}_p\left(4\right)+15{\hat{H}}_p\left(6\right)\left]\right\}$

208) utilize step 204) in the channel response value and the step 206 of the pilot sub-carrier position that obtains) in the channel response value of each symbol pilot sub-carrier position of obtaining, obtain the correction value w of the channel response value of each symbol.

Be w=[w ₁w ₂w ₃]

Wherein

$w_1=\frac{1}{2}[({\hat{H}}_p\left(1\right)-\hat{H}\left(2\right))+({\hat{H}}_p\left(2\right)-\hat{H}\left(11\right))]$

$w_2=\frac{1}{2}[({\hat{H}}_p\left(3\right)-\hat{H}\left(5\right))+({\hat{H}}_p\left(4\right)-\hat{H}\left(14\right))]$

$w_3=\frac{1}{2}[({\hat{H}}_p\left(5\right)-\hat{H}\left(8\right))+({\hat{H}}_p\left(6\right)-\hat{H}\left(17\right))]$

Step 210) channel response value of all data carriers of each symbol of each data block is revised, obtained the channel response value of all carrier waves.

For the 1st symbol, its channel response value is

For the 2nd symbol, its channel response value is

For the 3rd symbol, its channel response value is

Step 212) data that receive are carried out channel equalization, balanced data does

$D_{\mathrm{data},i}=\frac{D_{\mathrm{receive},i}}{{\hat{H}}_i}=(\frac{D_{\mathrm{receive},i}\left(1\right)}{{\hat{H}}_i\left(1\right)},\frac{D_{\mathrm{receive},i}\left(2\right)}{{\hat{H}}_i\left(2\right)},......\frac{D_{\mathrm{receive},i}\left(N\right)}{{\hat{H}}_i\left(N\right)})$

Wherein, i is call sign index i=1 in the data block, 2,3, and D _{Receive, i}N data subcarrier for i symbol receiving.

Step 214) the reception data after the channel equalization is carried out coherent demodulation.

The simulating scenes of the embodiment of the invention is: carrier frequency is 2.5GHz, and bandwidth is 10M, and sub-carrier number is 1024; Circulating prefix-length is 128 sampled points, and sub-carrier distribution manner is continuous subcarrier allocation, and number of subchannels is 48; Every frame has 15 symbols, and reception and number of transmit antennas are 1, system's ideal synchronisation; Modulation system is the QPSK modulation, and coded system is the CTC coding, and code check is 1/2.Channel circumstance is ITU VA 120Km/h.

Like Fig. 3; It is the comparison diagram of the channel estimating performance of first embodiment of the invention and traditional LS algorithm; The as can be seen from the figure bit error rate performance comparable situation of the inventive method (solid line) and traditional LS algorithm (dotted line); When the error rate was 1.0e-4, the signal to noise ratio of wanting required for the present invention was than the low 4dB of LS estimator.

As shown in Figure 4, the data block that second embodiment of the invention is used takies 7 symbol times on time domain, on frequency domain, take 12 carrier spacings.4 pilot sub-carriers and 80 data subcarriers are arranged in each data block.The 6th of first symbol and the 12nd carrier position are pilot sub-carrier, and are labeled as P _Recieve(1) and P _Recieve(2); The 3rd of the 5th symbol and the 9th carrier position are pilot sub-carrier, and are labeled as P _Recieve(3) and P _Recieve(4).

The flow chart of channel estimation methods according to the present invention, the supposing the system ideal synchronisation, the concrete steps of channel estimating are following:

202) data that receive (comprising pilot tone and data) are divided into data block.Wherein, each data block takies 7 symbol times on time domain, on frequency domain, takies 12 carrier spacings.The method of salary distribution of pilot sub-carrier and data subcarrier is as shown in Figure 4;

204) pilot sub-carrier of each data block is utilized the channel response value of LS algorithm estimating pilot frequency sub-carrier positions, promptly

${\hat{H}}_p=\frac{P_{\mathrm{receive}}}{P_{\mathrm{trans}}}=(\frac{P_{\mathrm{receive}}\left(1\right)}{P_{\mathrm{trans}}\left(1\right)},\frac{P_{\mathrm{receive}}\left(2\right)}{P_{\mathrm{trans}}\left(2\right)},\frac{P_{\mathrm{receive}}\left(3\right)}{P_{\mathrm{trans}}\left(3\right)},\frac{P_{\mathrm{receive}}\left(4\right)}{P_{\mathrm{trans}}\left(4\right)})$

Wherein, P _TransThe pilot frequency sequence that be to send is base station and travelling carriage known sequences all.

206) utilize the LS algorithm to obtain the channel response value of all carrier positions of whole data block to the channel response value

of the pilot sub-carrier position estimating to obtain.

H (k) is the channel frequency response at subcarrier k place, and 1≤k≤12 make p=[3,6,9,12], and P is the location index collection of pilot sub-carrier in the data block.H (k) is by estimating like drag:

$\hat{H}\left(k\right)=c_{1}k+c_2=C^H{q}_k=q_k^T{C}^{*}$

Wherein: $q_k^T=[k,1]$

C ^H＝[c ₁，c ₂]

The principle of confirming C is: $\underset{c}{\mathrm{Min}}\{{|{\hat{H}}_p\left(1\right)-C^H{q}_6|}^2+{|{\hat{H}}_p\left(2\right)-C^H{q}_{12}|}^2+{|{\hat{H}}_p\left(3\right)-C^H{q}_3|}^2+{|{\hat{H}}_p\left(4\right)-C^H{q}_9|}^2\}$

Finally calculate:

$\hat{H}\left(k\right)=k\left\{\frac{1}{630}\right[24{\hat{H}}_p\left(1\right)+18{\hat{H}}_p\left(2\right)+27{\hat{H}}_p\left(3\right)+{21\hat{H}}_p\left(4\right)\left]\right\}$

$+\left\{\frac{1}{630}\right[-90{\hat{H}}_p\left(1\right)+90{\hat{H}}_p\left(2\right)-18{0\hat{H}}_p\left(3\right)\left]\right\}$

208) utilize step 204) in the channel response value and the step 206 of the pilot sub-carrier position that obtains) in the channel response value of each symbol pilot sub-carrier position of obtaining, obtain the correction value w of the channel response value of each symbol.

Be w=[w ₁w ₂w ₃w ₄w ₅w ₆w ₇]

Wherein

$w_1=\frac{1}{2}[({\hat{H}}_p\left(1\right)-\hat{H}\left(6\right))+({\hat{H}}_p\left(2\right)-\hat{H}\left(12\right))]$

$w_5=\frac{1}{2}[({\hat{H}}_p\left(3\right)-\hat{H}\left(3\right))+({\hat{H}}_p\left(4\right)-\hat{H}\left(9\right))]$

$w_2=\frac{3}{4}{w}_1+\frac{1}{4}{w}_5$

$w_3=\frac{1}{2}{w}_1+\frac{1}{2}{w}_5$

$w_4=\frac{1}{4}{w}_1+\frac{3}{4}{w}_5$

$w_6=-\frac{1}{4}{w}_1+\frac{5}{4}{w}_5$

$w_7=-\frac{1}{2}{w}_1+\frac{3}{2}{w}_5$

210) channel response value of all data carriers of each symbol of each data block is revised, obtained the channel response value of all carrier waves.

For the 1st symbol, its channel response value is

For the 2nd symbol, its channel response value is

For the 3rd symbol, its channel response value is

……………………

For the 7th symbol, its channel response value is

212) data that receive are carried out channel equalization, balanced data does

$D_{\mathrm{data},i}=\frac{D_{\mathrm{receive},i}}{{\hat{H}}_i}=(\frac{D_{\mathrm{receive},i}\left(1\right)}{{\hat{H}}_i\left(1\right)},\frac{D_{\mathrm{receive},i}\left(2\right)}{{\hat{H}}_i\left(2\right)},......\frac{D_{\mathrm{receive},i}\left(N\right)}{{\hat{H}}_i\left(N\right)})$

Wherein, i is call sign index i=1 in the data block, 2,3 ... 7, D _{Receive, i}N data subcarrier for i symbol receiving.

214) the reception data after the channel equalization are carried out coherent demodulation.

Though explained with reference to the particular step of implementing with particular order and shown method disclosed herein; But should be appreciated that; Can make up, segment these steps or resequence and form a kind of equivalent method; This does not deviate from teaching content of the present invention, therefore, and under the situation that does not deviate from spirit of the present invention and essence thereof; Those of ordinary skill in the art work as can make various corresponding changes and distortion according to the present invention, but these corresponding changes and distortion all should belong to the protection range of the appended claim of the present invention.

Claims (10)

1. the method for channel estimating in the ofdm system is characterized in that, may further comprise the steps:

1.1) continuous data that receives is divided into some time-frequency two-dimensional data blocks;

1.2) utilize pilot tone to obtain the channel estimation value that channel estimation values of pilot frequency positions also further obtains all carrier frequency positions of said two-dimensional blocks of data;

1.3) obtain the time domain correction value according to the difference of the channel estimation value of the carrier frequency position of channel estimation values of pilot frequency positions and this pilot tone place carrier wave;

1.4) add the channel estimation value that the above time domain correction value is obtained all data in the said data block according to the channel estimation value of said all carrier frequency positions.

2. according to the method for the said channel estimating of claim 1, it is characterized in that each said time-frequency two-dimensional data blocks takies m symbol in time domain, take n carrier wave on the frequency domain, comprise m * n data, wherein: m, n are natural number.

3. according to the method for the said channel estimating of claim 1, it is characterized in that said step 1.2) in carry out the channel estimating of said pilot frequency locations with the least-squares estimation method.

4. according to the method for the said channel estimating of claim 3, it is characterized in that said channel estimation values of pilot frequency positions

According to formula

Obtain, wherein: P _TransBe the pilot frequency sequence that sends, P _ReceiveBe the pilot frequency sequence that receives, p is the corresponding time domain sequence number of pilot tone, and k ' is the corresponding frequency domain sequence number of pilot tone.

5. according to the method for the said channel estimating of claim 3, it is characterized in that said step 1.2) in further adopt the least-squares estimation method to carry out the channel estimating of said all carrier frequency positions.

6. according to the method for the said channel estimating of claim 5; It is characterized in that; The channel estimation value of all carrier frequency positions of said two-dimensional blocks of data

is according to obtaining with drag; H (k) is the channel frequency response at subcarrier k place; K is the corresponding frequency domain sequence numbers of each data of data block, and P representes the location index collection of pilot sub-carrier in the data block:

$\hat{H}\left(k\right)=c_{1}k+c_2=C^H{q}_k=q_k^T{C}^{*}$

Wherein: $q_k^T=[k,1]$

C ^H＝[c ₁，c ₂]

The principle of confirming C is: $\underset{c}{\mathrm{Min}}\left\{\underset{k\&Element;P}{\mathrm{\&Sigma;}}{|H\left(k\right)-C^H{q}_k|}^2\right\}.$

7. according to the method for the said channel estimating of claim 1, it is characterized in that one or more pilot tones are all arranged on the carrier wave of arbitrary time slot of said data block, have the corresponding time domain correction value of the time slot w of an above pilot tone _iGet the average of said difference.

8. according to the method for the said channel estimating of claim 1, it is characterized in that i symbol is at the channel estimation value at subcarrier k place

According to formula

Obtain, wherein: i is the corresponding time domain sequence numbers of each data of data block, and k is the corresponding frequency domain sequence numbers of each data of data block,

Be the channel estimation value at k place, carrier frequency position, w _iThe correction value of the channel response value of i symbol on the expression time domain.

9. according to the method for the said channel estimating of claim 1; It is characterized in that; This method comprises that also the channel response value of utilizing all data carries out channel equalization to receiving data block, and then balanced data is carried out coherent demodulation.

10. according to the method for the said channel estimating of claim 1, it is characterized in that said ofdm system is based on the ofdm system of IEEE802.16a/802.16d/802.16e/LTE standard.

Images