CN100405790C - Self-adaptive channel estimation method based on least squares criterion of two-dimensional iteration - Google Patents
Self-adaptive channel estimation method based on least squares criterion of two-dimensional iteration Download PDFInfo
- Publication number
- CN100405790C CN100405790C CNB200410049742XA CN200410049742A CN100405790C CN 100405790 C CN100405790 C CN 100405790C CN B200410049742X A CNB200410049742X A CN B200410049742XA CN 200410049742 A CN200410049742 A CN 200410049742A CN 100405790 C CN100405790 C CN 100405790C
- Authority
- CN
- China
- Prior art keywords
- channel
- value
- channel estimation
- matrix
- channel estimator
- Prior art date
- Legal status (The legal status is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the status listed.)
- Expired - Fee Related
Links
Images
Abstract
The present invention relates to a self-adaptive channel estimation method based on a least squares criterion of two-dimensional iteration. The method adopts a 2D-RLS criterion, namely that the self-adaptive channel estimation is carried out by making use of the two-dimensional correlativity of the channels, and any channel statistical property does not need to be known. The present invention can effectively increase the accuracy of the channel estimation results, solve the inaccurate problem of the channel estimation results caused by that the statistical property of the channels can not be accurately obtained, or the two-dimensional correlativity of the channels can not be sufficiently utilized. The present invention has the other advantage that the split processing between the computation complexity and the channel estimation performance can be conveniently realized by selecting the size of input vectors of a channel estimation device of signals, and the physical structure of the channel estimation device does not need to be changed.
Description
Technical field
The present invention relates to communication technical field, relate in particular in a kind of ofdm system adaptive channel estimation method based on 2D-RLS (two-dimentional iterative least square) criterion.
Background technology
Along with the high speed development of Internet (the Internet) and mobile communication technology, there is potential great demand in wireless data service at a high speed.Yet it is very difficult that high-speed data service is provided in abominable wireless channel environment, and therefore, industry is always for providing the Radio Transmission Technology with superperformance to carry out effort.Existing OFDM (OFDM) technology has the ability of antagonism ISI (intersymbol interference), can provide very high spectrum efficiency simultaneously, therefore is regarded as the professional most possible Radio Transmission Technology that adopts of high speed wireless data.OFDM has application very widely, comprising: XDSL (Digital Subscriber Loop), DAB/DVB (digital audio/video broadcasting), WLAN standard IEEE 802.11a and HIPERLAN/2, wireless metropolitan area network standard IEEE 802.16, or the like.
In order to guarantee that communication system can have good performance in wireless channel environment, the wireless fading channel that becomes in the time of must be to multipath is estimated.Can think the order of accuarcy of channel estimating has determined to a great extent whether system can provide good wireless transmission quality.Promptly in ofdm system, the quality of channel estimating plays key effect to the performance of ofdm system.At present the channel estimation methods that adopts roughly can be divided into two big classes: blind estimation and based on the channel estimating of pilot tone.Though described blind estimation does not have the higher availability of frequency spectrum because do not need pilot tone, complexity is very high and performance is bad, at present still can't be practical; Described channel estimating based on pilot tone is divided into again based on LS (least square) criterion with based on MMSE (least mean-square error) criterion.Though the LS channel estimating is simple, compare with the MMSE channel estimating, in order to reach identical channel estimating performance (MSE (mean square error) with channel estimating weighs), there is SNR (signal to noise ratio) loss of 10-15dB.But in order to realize the MMSE channel estimating, need know channel statistical characteristic accurately, this can't realize in practice.
Document (Y. (G.) Li is arranged, L.J.Cimini, and N.R.Sollenberger, " Robustchannel estimation for OFDM systems with rapid dispersive fadingchannels; " IEEE Trans.Commun.vol.46, pp.902-915, July 1998.) the Robust channel estimating proposed, can utilize the two-dimensional correlation characteristic of channel, and insensitive to the variation of channel statistical characteristic.But the Robust channel estimating still needs to know in advance several channel parameters: maximum doppler frequency, maximum multipath time delay and noise power, clearly these three parameters are difficult to obtain.So in order to adapt to unknown channel as much as possible, the Robust channel estimating need suppose that maximum doppler frequency, maximum multipath time delay and noise power all get bigger value, this has inevitably caused the accuracy of channel estimating to descend.
In addition, also has document (D.Schafhuber, G.Matz, and F.Hlawatsch, " AdaptiveWiener filters for time-varying channel estimation in wireless OFDMsystems ", ICASSP ' 03, vol.4, pp.IV-688-IV-691,6-10Apr.2003.) channel estimating based on time domain RLS (iterative least square) criterion proposed, though this method does not need to know channel statistic property, can only utilize the correlation of channel on this dimension of time-domain.Simultaneously, very sensitive based on the channel estimating of time domain RLS criterion for the variation of channel multi-path time delay, and the variation of this channel multi-path time delay is very common in actual environment.
Summary of the invention
In view of above-mentioned existing in prior technology problem, the purpose of this invention is to provide a kind of adaptive channel estimation method, thereby improve the accuracy of channel estimating in the wireless communication system based on two-dimentional iterative least square criterion, improve the performance of wireless communication system.
The objective of the invention is to be achieved through the following technical solutions:
The invention provides a kind of adaptive channel estimation method, comprising based on two-dimentional iterative least square criterion:
A, when orthogonal frequency division multiplex OFDM system is started working, transmitting terminal need send the known OFDM symbol of setting quantity and be set to complete 0 matrix as the initial value of the coefficient matrix of training sequence and channel estimator, receiving terminal receives the OFDM symbol of described setting quantity, the pairing frequency domain channel value of OFDM (OFDM) symbol that receives is rearranged, the described L of being rearranged for has and to be in the individual vector that to be arranged in a dimension be LM * 1 of the LM on the same sub-carrier location in each OFDM symbol in the pairing frequency domain channel value of OFDM symbol of N subcarrier, or be arranged in a plurality of vectors, and the dimension of dimension when only being a vector that a plurality of vectors add together is identical, promptly also be LM * 1, M=N/2
z, and 2
zUsually should be less than or equal to log
2N, and z is a positive integer;
B, will multiply each other through the frequency domain channel value that rearranges and last one constantly the associate matrix of channel estimator coefficient matrix with each self-corresponding channel estimator, the value of Huo Deing merges afterwards as channel estimation results separately.。
Described steps A comprises:
Receiving terminal use through data after the demodulate/decode and the output of former FFT (fast Fourier transform) module through decline and the signal that has a noise according to LS (least square) criterion carry out decision-feedback handle obtain channel estimator needs but still without the input signal that rearranges;
Described decision-feedback handled be in the individual vector that to be arranged in a dimension be LM * 1 of the LM on the same sub-carrier location in each OFDM symbol in the pairing frequency domain channel value of OFDM symbol that L of obtaining have N subcarrier, or be arranged in a plurality of vectors, and the dimension of dimension when only being a vector that a plurality of vectors add together is identical, promptly also be LM * 1, M=N/2
z, and 2
zUsually should be less than or equal to log
2N, and z is a positive integer.
Described step B comprises:
One or more total dimensions vector that is LM * 1 is multiplied each other with last one constantly the associate matrix of channel estimator coefficient matrix of each self-corresponding channel estimator respectively, and the value of Huo Deing merges afterwards as channel estimation results separately.
Described channel estimator is 2D-RLS (two-dimentional iterative least square) channel estimator, and described two dimension is the time dimension and the frequency dimension of frequency domain channel.
The described processing procedure that also comprises the coefficient matrix that obtains described channel estimator based on the adaptive channel estimation method of two-dimentional iterative least square criterion:
C, calculate the difference between current channel estimation results and the ideal communication channel value, and calculate the gain vector value of the coefficient matrix renewal amount of current adjustment channel estimator according to current channel estimator input vector and a last moment inverse matrix of autocorrelation matrix thereof and previously selected forgetting factor;
D, carry out the adjustment of the coefficient matrix of channel estimator according to the mode of setting according to described difference and current gain vector value, obtain the coefficient matrix of the channel estimator used in the channel estimation process next time, inverse matrix of autocorrelation matrix and previously selected forgetting factor and current gain vector value are upgraded the inverse matrix of autocorrelation matrix of channel estimator input vector according to the mode of setting constantly according to current channel estimator input vector and last simultaneously.
Among the present invention, the acquisition process of the described ideal communication channel value of step C comprises:
Process decline that data after receiving terminal uses demodulate/decode and former FFT (fast Fourier transform) module are exported and the signal that has noise are estimated frequency domain channel according to LS (least square) criterion;
Result's process IDFT (contrary discrete fourier transform) processing of estimating to obtain is transformed to time-domain signal;
Described time-domain signal is carried out most powerful path catches to handle and obtain wherein stronger time-domain signal sample value, and time-domain signal sample value that will be stronger through DFT (discrete fourier transform) conversion return frequency domain as the ideal communication channel value.
In the described adaptive channel estimation method based on two-dimentional iterative least square criterion:
Described step C comprises:
Calculate the difference ξ (n) between current channel estimation results and the ideal communication channel value;
And according to current channel estimator input vector and last constantly inverse matrix of autocorrelation matrix and previously selected forgetting factor calculate the gain vector value k (n) of the coefficient matrix renewal amount of current adjustment channel estimator, it is Q (n-1) p (n)/{ λ+p that the iterative computation of described k (n) is closed
H(n) Q (n-1) p (n) }, L the pairing LM of the OFDM symbol vector that the frequency domain channel value is constituted that p (n) rearranges for current process, λ is a forgetting factor, value is the arithmetic number less than 1, channel variation should be worth big more more slowly, Q (n) is the inverse matrix of autocorrelation matrix of p (n), and it is λ that the iterative computation of Q (n) is closed
-1Q (n-1)-λ
-1K (n) p
H(n) Q (n-1), p
H(n) be the conjugate transpose of p (n), described n is then for carrying out the iterations count value of channel estimating;
Described step D comprises:
Determine the value of the coefficient matrix G (n) of the channel estimator used in the channel estimation process next time according to described difference ξ (n) and current gain vector value k (n), described G (n)=G (n-1)+k (n) * ξ
H(n), ξ
H(n) be the conjugate transpose of ξ (n).
Described steps A comprises:
When ofdm system is started working, transmitting terminal need send the known OFDM symbol of setting quantity as training sequence, receiving terminal receives the OFDM symbol of described setting quantity, and carry out channel estimating according to LS (least square) criterion, obtain the frequency domain channel value of the OFDM symbol correspondence of setting quantity, as the initial input value of channel estimator;
The vector that it is LM * 1 that the frequency domain channel value that L of initially receiving is had a pairing same sub-carrier location in the OFDM symbol of N subcarrier is arranged in one or more total dimensions, M=N/2
z, and z is a positive integer.
In described adaptive channel estimation method based on two-dimentional iterative least square criterion, when ofdm system is started working, the initial value that carries out the coefficient matrix of the channel estimator described in the channel estimation process at described receiving terminal is complete 0 matrix, the initial value of the inverse matrix of autocorrelation matrix of the input vector of described channel estimator is that the inverse of regularization parameter multiply by unit matrix, described regularization parameter is an arithmetic number, and SNR (signal to noise ratio) is high more, and the value of being somebody's turn to do is more little.
Method of the present invention is applicable to ofdm system, comprises MIMO-OFDM (many antennas OFDM) system.
As seen from the above technical solution provided by the invention, the present invention is owing to adopted the 2D-RLS criterion, promptly utilize the two-dimensional correlation of channel to carry out adaptive channel estimating, need not to know any channel statistical characteristic, can improve the accuracy of channel estimation results effectively, solve in the prior art owing to can't accurately obtaining the channel estimation results inaccurate problem that two-dimensional correlation that the channel statistical characteristic maybe can't make full use of channel causes.
Simultaneously, the adaptive channel estimation method based on the 2D-RLS criterion of the present invention only need several OFDM symbols time just very rapid convergence to stable state.Another one advantage of the present invention be can be easily the compromise that is implemented between computation complexity and the channel estimating performance of the size of input vector by the selective channel estimator handle, and need not to change the physical structure of channel estimator.
Evidence method of the present invention will obviously be better than having other channel estimation methods that writes down in the document now on performance, and goes for the ofdm system under the different radio channel condition.
Description of drawings
Fig. 1 is the structural representation of ofdm system;
Fig. 2 is the rearrangement process schematic diagram of channel estimator input vector p;
Fig. 3 is the theory diagram of 2D-RLS algorithm among the present invention;
Fig. 4 is the structural representation of parallel 2D-RLS channel estimator;
Fig. 5 is the production process schematic diagram of reference signal;
The OFDM symbol frequency domain channel value that Fig. 6 is suitable for for the present invention choose the pattern schematic diagram;
Fig. 7 chooses the pattern schematic diagram for the inapplicable OFDM symbol of the present invention frequency domain channel value.
Embodiment
Method of the present invention is a kind of adaptive channel estimation method based on the 2D-RLS criterion, for realizing the present invention, the known OFDM symbol that needed to send earlier some (be generally 5 get final product with interior) before ofdm system begins to carry out channel estimating is as training sequence, just can enter a kind of then is the channel estimation process that iterates of chronomere with the OFDM symbol, and each iteration has comprised three steps altogether: channel estimating; Calculate the inverse matrix of autocorrelation matrix of the input vector of evaluated error, gain vector and channel estimator; The adaptive adjustment of carrying out the channel estimator coefficient matrix.And carry out in the process of transfer of data at ofdm system, no longer need to insert pilot tone or training symbol, have higher spectrum efficiency.
Now in conjunction with the accompanying drawings to the embodiment of method of the present invention:
In ofdm system, be provided with the channel estimation process part at receiving terminal, the particular location of channel estimation process part is in FFT (fast Fourier transform) module and constellation and separates between the mapping block referring to Fig. 1.
The specific implementation process of channel estimation methods of the present invention is described below referring to accompanying drawing:
Step 1: when ofdm system initially enters operating state, in order to use the 2D-RLS algorithm to carry out channel estimating, at first need transmitting terminal to send the long training sequence of L OFDM symbol that is, L is usually less than 5, so that receiving terminal can carry out initial channel estimating according to LS (least square) criterion under initial situation, obtain corresponding frequency domain channel value;
The sub-carrier number of supposing ofdm system is N, simultaneously, also needs the initial value of the coefficient matrix of channel estimator to be set to G (0)=0, and promptly G (0) is complete 0 matrix.
Step 2: receiving terminal receives training sequence and carries out initial channel estimating according to LS (least square) criterion, obtain corresponding frequency domain channel value after, it is rearranged the actual initial input value that obtains channel estimator;
Column vector p (n) is the input of 2D-RLS channel estimator, n is the iterations count value of carrying out channel estimating, is rearranged by LM frequency domain channel value in L the OFDM symbol before the current time to obtain, as shown in Figure 2, dimension is LM * 1, wherein M=N/2
z, z is positive integer such as value 0,1,2 as required, but should satisfy 2 usually
zBe less than or equal to log
2N, the value size of M has determined the size of the quantity chosen from the frequency domain channel value of N subcarrier correspondence;
In addition, usually when the number of subcarriers of ofdm system hour, use a channel estimator that the pairing frequency domain channel of current OFDM symbol is estimated to get final product.And when the number of subcarriers of ofdm system is very big, then LM frequency domain channel value in L the OFDM symbol can be arranged as the identical vector of a plurality of dimensions, and multiply each other with last one constantly the conjugate transpose of coefficient matrix of each self-corresponding channel estimator respectively, obtain the channel estimating of the different piece of the pairing frequency domain channel of current OFDM symbol respectively, promptly when the sub-carrier number of ofdm system is very big, in order to reduce computation complexity, improve the real-time of algorithm, subcarrier in the OFDM symbol can be divided into identical several groups of size, use parallel 2D-RLS channel estimator to carry out channel estimating then, as shown in Figure 4, just obtain the channel estimation results of current OFDM symbol after the combination of the channel estimation results of each channel estimator.
Step 3: carry out channel estimating according to the coefficient matrix of channel estimator and through L the pairing LM of an OFDM symbol frequency domain channel value that rearranges, obtain current channel estimation results h (1)=G
H(0) * and p (1), certainly, corresponding channel estimation results then is h (n)=G in the subsequent communication channel estimation procedure
H(n-1) * p (n);
Wherein, column vector h (n) is the frequency domain channel value that needs the current OFDM symbol of estimation, and dimension is N * 1; Matrix G (n) is the coefficient matrix of 2D-RLS channel estimator, and dimension is LM * N, G
H(n) be the associate matrix of G (n).
For guarantee the channel estimation process process continue carry out, after each channel estimation process process finishes, also need the coefficient matrix of described channel estimator is adjusted, coefficient matrix with the channel estimator after the acquisition renewal, and be used for next time channel estimation process process, referring to Fig. 3, the processing procedure that the coefficient matrix of described channel estimator is adjusted is as follows:
Step 4: calculate the error between current channel estimation results and ideal communication channel value;
For calculating channel evaluated error ξ (n), will be appreciated that ideal channel response h ' theoretically (n), h ' (n) can't obtain in real system, therefore can only use corresponding estimated value; The present invention obtains h ' processing procedure (n) as shown in Figure 5, at first use data after the demodulate/decode (i.e. decoding) that receiving terminal receives and received signal to carry out DD (decision-feedback) and obtain frequency domain rough estimate value according to LS (least square) criterion, use IDFT (contrary discrete fourier transform) to handle this frequency domain rough estimate value transform then to time domain, re-use the most powerful path harvesting policy and obtain several more intense in time-domain signal footpaths, remaining time domain channel sample value all is changed to 0, and then use DFT (discrete fourier transform) processing conversion and return frequency domain, obtain the ideal channel response h ' estimated value (n) of more accurate current ofdm system;
Obtain described h ' (n) just can calculate after the value current channel estimation errors ξ (n)=h ' (n)-h (n).
Step 5: simultaneously for the coefficient matrix of channel estimator is adjusted, also need to calculate the corresponding gain vector of acquisition, described gain vector is to calculate acquisition according to current channel estimator input vector and last an inverse matrix of autocorrelation matrix and previously selected forgetting factor constantly;
Described gain vector k (n) is used to adjust the renewal amount of the coefficient matrix of channel estimator, and the iteration of described k (n) more new relation be Q (n-1) p (n)/{ λ+p
H(n) Q (n-1) p (n) }, L the pairing LM of the OFDM symbol vector that the frequency domain channel value is constituted that p (n) rearranges for current process;
Q (n) is the inverse matrix of autocorrelation matrix of current channel estimator input vector p (n); Initial value Q (0)=δ of Q (n)
-1I, wherein δ is the regularization parameter, and value is an arithmetic number, and choosing of this value is relevant with the SNR (signal to noise ratio) of channel, and SNR is big more, and the value of being somebody's turn to do is more little, and I is a unit matrix; At the Q (n) described in the follow-up computational process then is to obtain by iterative computation, and its iteration more new relation is λ
- 1Q (n-1)-λ
-1K (n) p
H(n) Q (n-1), p
H(n) be the conjugate transpose of p (n), described n is then still for carrying out the iterations count value of channel estimating;
Described forgetting factor λ is generally less than 1 but approaches 1 arithmetic number, and concrete value is relevant with channel variance situation, and this is worth big more channel variation more slowly.
Step 6: obtained described k (n) and ξ (n), just can adjust the coefficient matrix of channel estimator by adaptive channel estimator coefficient matrix update mechanism, be that the adaptive adjustment of carrying out the coefficient matrix of channel estimator according to k (n) and ξ (n) value is upgraded, to obtain to be used for the coefficient matrix of the channel estimator of channel estimating next time, referring to Fig. 3, this step is finished by the coefficient matrix processing section of adaptive updates channel estimator;
Be specially, next time the coefficient matrix G (n) of the channel estimator of using in the channel estimation process=G (n-1)+k (n) * ξ
H(n);
Step 7: after the adjustment of finishing the coefficient matrix of described channel estimator is upgraded,, also need upgrade the inverse matrix of autocorrelation matrix of channel estimator input vector according to the description in the step 5 for guarantee to adjust the needs of update calculation next time.
Generally, choosing of regularization parameter δ and forgetting factor λ has very big influence to the performance based on the algorithm of RLS criterion, but found through experiments based on the performance of the adaptive channel estimation method of 2D-RLS criterion very insensitive for choosing of regularization parameter δ and forgetting factor λ.For the value of most of regularization parameter δ and forgetting factor λ, the present invention only needs the time of several OFDM symbols just can converge to stable status.
Among the present invention, only need to send L known OFDM symbol as training sequence in the training stage (being that ofdm system is started working the stage), and no longer needing to insert pilot tone or training symbol at the data transfer phase of ofdm system, this moment, the input signal p (n) of channel estimator carried out rearranging acquisition again behind the DD (decision-feedback) according to the LS criterion by data and received signal after the use demodulate/decode.
Carry out in the normal data transmission procedure at ofdm system, method of the present invention mainly comprises three processing procedures: (1) determined according to the coefficient matrix G (n-1) of the channel estimator in a last moment and the input signal p (n) of channel estimator that current channel estimation results was h (n)=G
H(n-1) * p (n); The gain vector k (n) of (2) the difference ξ (n) that calculates between current channel estimation results and ideal communication channel value with reference to above-mentioned steps 4, and calculating current time; (3) carry out the adjustment of the coefficient matrix of channel estimator with reference to above-mentioned steps 6, and calculate the inverse matrix of autocorrelation matrix Q (n) of the input vector of current time with reference to above-mentioned steps 5,7.(1), (2) are carried out in circulation in ofdm system, (3) three processes just can realize the real-time estimation to channel.
Among the present invention, the problem that should be noted that is, for the result who keeps each channel estimating can accumulate, promptly make channel estimating to carry out with adaptive iterative manner, this method has requirement for the position of LM frequency domain channel value in L the OFDM symbol that is used for constituting p (n), being specially what require to choose is identical sub-carrier positions in each OFDM symbol, the frequency domain channel value of promptly choosing OFDM symbol as shown in Figure 6 is satisfactory, and the frequency domain channel value of choosing OFDM symbol as shown in Figure 7 then is undesirable.
Among the present invention, by being different emitting antenna selecting suitable training sequences, the adaptive channel estimation method based on the 2D-RLS criterion provided by the invention also is suitable for for MIMO-OFDM (many antennas OFDM) system.Described is that the concrete grammar of different emitting antenna selecting suitable training sequence can reference literature I.Barhumi, G.Leus, and M.Moonen, " Optimal TrainingDesign for MIMO OFDM Systems in Mobile Wireless Channels ", IEEETrans.Signal Processing, vol.51, pp.1615-1624, the training sequence design that June 2003. provides.
The above; only for the preferable embodiment of the present invention, but protection scope of the present invention is not limited thereto, and anyly is familiar with those skilled in the art in the technical scope that the present invention discloses; the variation that can expect easily or replacement all should be encompassed within protection scope of the present invention.Therefore, protection scope of the present invention should be as the criterion with the protection range of claims.
Claims (10)
1. adaptive channel estimation method based on two-dimentional iterative least square criterion is characterized in that comprising:
A, when orthogonal frequency division multiplex OFDM system is started working, transmitting terminal need send the known OFDM symbol of setting quantity and be set to complete 0 matrix as the initial value of the coefficient matrix of training sequence and channel estimator, receiving terminal receives the OFDM symbol of described setting quantity, the pairing frequency domain channel value of the orthogonal frequency division multiplex OFDM symbol that receives is rearranged, the described L of being rearranged for has and to be in the individual vector that to be arranged in a dimension be LM * 1 of the LM on the same sub-carrier location in each OFDM symbol in the pairing frequency domain channel value of OFDM symbol of N subcarrier, or be arranged in a plurality of vectors, and the dimension of dimension when only being a vector that a plurality of vectors add together is identical, promptly also be LM * 1, M=N/2
z, and 2
zShould be less than or equal to log
2N, and z is a positive integer;
B, will multiply each other through last one constantly the associate matrix of channel estimator coefficient matrix of the frequency domain channel value that rearranges and each self-corresponding channel estimator, the value of Huo Deing merges afterwards as channel estimation results separately.
2. the adaptive channel estimation method based on two-dimentional iterative least square criterion according to claim 1 is characterized in that described steps A comprises:
Receiving terminal use through data after the demodulate/decode and the output of former fast Fourier transform FFT module through decline and the signal that has a noise according to least square LS criterion carry out decision-feedback handle obtain channel estimator needs but still without the input signal that rearranges;
Described decision-feedback handled be in the individual vector that to be arranged in a dimension be LM * 1 of the LM on the same sub-carrier location in each OFDM symbol in the pairing frequency domain channel value of OFDM symbol that L of obtaining have N subcarrier, or be arranged in a plurality of vectors, and the dimension of dimension when only being a vector that a plurality of vectors add together is identical, promptly also be LM * 1, M=N/2
z, and 2
zShould be less than or equal to log
2N, and z is a positive integer.
3. the adaptive channel estimation method based on two-dimentional iterative least square criterion according to claim 2 is characterized in that described step B comprises:
One or more total dimensions vector that is LM * 1 is multiplied each other with last one constantly the associate matrix of channel estimator coefficient matrix of each self-corresponding channel estimator respectively, and the value of Huo Deing merges afterwards as channel estimation results separately.
4. the adaptive channel estimation method based on two-dimentional iterative least square criterion according to claim 1, it is characterized in that described channel estimator is two-dimentional iterative least square 2D-RLS channel estimator, and described two dimension is the time dimension and the frequency dimension of frequency domain channel.
5. according to claim 1,2,3 or 4 described adaptive channel estimation methods, it is characterized in that this method also comprises the processing procedure of the coefficient matrix that obtains described channel estimator based on two-dimentional iterative least square criterion:
C, calculate the difference between current channel estimation results and the ideal communication channel value, and calculate the gain vector value of the coefficient matrix renewal amount of current adjustment channel estimator according to current channel estimator input vector and a last moment inverse matrix of autocorrelation matrix thereof and previously selected forgetting factor;
D, carry out the adjustment of the coefficient matrix of channel estimator according to the mode of setting according to described difference and current gain vector value, obtain the coefficient matrix of the channel estimator used in the channel estimation process next time, inverse matrix of autocorrelation matrix and previously selected forgetting factor and current gain vector value are upgraded the inverse matrix of autocorrelation matrix of channel estimator input vector according to the mode of setting constantly according to current channel estimator input vector and last simultaneously.
6. the adaptive channel estimation method based on two-dimentional iterative least square criterion according to claim 5 is characterized in that the acquisition process of the described ideal communication channel value of step C comprises:
Process decline that data after receiving terminal uses demodulate/decode and former fast Fourier transform FFT module are exported and the signal that has noise are estimated frequency domain channel according to least square LS criterion;
The result who estimates to obtain is transformed to time-domain signal through contrary discrete fourier transform IDFT processing;
Described time-domain signal is carried out most powerful path catches to handle and obtain wherein stronger time-domain signal sample value, and time-domain signal sample value that will be stronger through discrete fourier transform DFT conversion return frequency domain as the ideal communication channel value.
7. the adaptive channel estimation method based on two-dimentional iterative least square criterion according to claim 6 is characterized in that:
Described step C comprises:
Calculate the difference ξ (n) between current channel estimation results and the ideal communication channel value;
And according to current channel estimator input vector and last constantly inverse matrix of autocorrelation matrix and previously selected forgetting factor calculate the gain vector value k (n) of the coefficient matrix renewal amount of current adjustment channel estimator, it is Q (n-1) p (n)/{ λ+p that the iterative computation of described k (n) is closed
H(n) Q (n-1) p (n) }, L the pairing LM of the OFDM symbol vector that the frequency domain channel value is constituted that p (n) rearranges for current process, λ is a forgetting factor, value is the arithmetic number less than 1, channel variation should be worth big more more slowly, Q (n) is the inverse matrix of autocorrelation matrix of p (n), and it is λ that the iterative computation of Q (n) is closed
-1Q (n-1)-λ
-1K (n) p
H(n) Q (n-1), p
H(n) be the conjugate transpose of p (n), described n is then for carrying out the iterations count value of channel estimating;
Described step D comprises:
Determine the value of the coefficient matrix G (n) of the channel estimator used in the channel estimation process next time according to described difference ξ (n) and current gain vector value k (n), described G (n)=G (n-1)+k (n) * ξ
H(n), ξ
H(n) be the conjugate transpose of ξ (n).
8. the adaptive channel estimation method based on two-dimentional iterative least square criterion according to claim 7 is characterized in that described steps A comprises:
When ofdm system is started working, transmitting terminal need send the known OFDM symbol of setting quantity as training sequence, receiving terminal receives the OFDM symbol of described setting quantity, and carry out channel estimating according to least square LS criterion, obtain the frequency domain channel value of the OFDM symbol correspondence of setting quantity, as the initial input value of channel estimator;
The vector that it is LM * 1 that the frequency domain channel value that L of initially receiving is had a pairing same sub-carrier location in the OFDM symbol of N subcarrier is arranged in one or more total dimensions, M=N/2
z, and z is a positive integer.
9. the adaptive channel estimation method based on two-dimentional iterative least square criterion according to claim 8, it is characterized in that, when ofdm system is started working, the initial value that carries out the coefficient matrix of the channel estimator described in the channel estimation process at described receiving terminal is complete 0 matrix, the initial value of the inverse matrix of autocorrelation matrix of the input vector of described channel estimator is that the inverse of regularization parameter multiply by unit matrix, described regularization parameter is an arithmetic number, and the high more value of being somebody's turn to do of signal to noise ratio snr is more little.
10. the adaptive channel estimation method based on two-dimentional iterative least square criterion according to claim 1 is characterized in that this method is applicable to ofdm system, comprises many antennas OFDM MIMO-OFDM system.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CNB200410049742XA CN100405790C (en) | 2004-06-25 | 2004-06-25 | Self-adaptive channel estimation method based on least squares criterion of two-dimensional iteration |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CNB200410049742XA CN100405790C (en) | 2004-06-25 | 2004-06-25 | Self-adaptive channel estimation method based on least squares criterion of two-dimensional iteration |
Publications (2)
| Publication Number | Publication Date |
|---|---|
| CN1595924A CN1595924A (en) | 2005-03-16 |
| CN100405790C true CN100405790C (en) | 2008-07-23 |
Family
ID=34665816
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CNB200410049742XA Expired - Fee Related CN100405790C (en) | 2004-06-25 | 2004-06-25 | Self-adaptive channel estimation method based on least squares criterion of two-dimensional iteration |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN100405790C (en) |
Families Citing this family (5)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN101399803B (en) * | 2007-09-27 | 2011-04-13 | 大唐移动通信设备有限公司 | Multi-user detection method and device for OFDM signal |
| CN101447969B (en) * | 2008-12-31 | 2011-04-20 | 宁波大学 | Channel estimation method of multi-band orthogonal frequency division multiplexing ultra wide band system |
| CN102065035B (en) * | 2008-12-31 | 2014-03-12 | 宁波大学 | Channel estimation method of multi-band orthogonal frequency-division multiplexing ultra-wideband system |
| CN109450830B (en) * | 2018-12-26 | 2021-09-21 | 重庆大学 | Channel estimation method based on deep learning in high-speed mobile environment |
| CN114567525B (en) * | 2022-01-14 | 2023-07-28 | 北京邮电大学 | Channel estimation method and device |
Citations (4)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| US20010036235A1 (en) * | 1999-12-22 | 2001-11-01 | Tamer Kadous | Channel estimation in a communication system |
| CN1437338A (en) * | 2003-03-08 | 2003-08-20 | 华中科技大学 | Channel estimation method for orthogonal frequency-division multiplexing communication system |
| CN1472967A (en) * | 2002-07-29 | 2004-02-04 | 矽统科技股份有限公司 | Wireless local network channel estimating devices |
| US20040086055A1 (en) * | 1998-12-31 | 2004-05-06 | Ye Li | Pilot-aided channel estimation for OFDM in wireless systems |
-
2004
- 2004-06-25 CN CNB200410049742XA patent/CN100405790C/en not_active Expired - Fee Related
Patent Citations (4)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| US20040086055A1 (en) * | 1998-12-31 | 2004-05-06 | Ye Li | Pilot-aided channel estimation for OFDM in wireless systems |
| US20010036235A1 (en) * | 1999-12-22 | 2001-11-01 | Tamer Kadous | Channel estimation in a communication system |
| CN1472967A (en) * | 2002-07-29 | 2004-02-04 | 矽统科技股份有限公司 | Wireless local network channel estimating devices |
| CN1437338A (en) * | 2003-03-08 | 2003-08-20 | 华中科技大学 | Channel estimation method for orthogonal frequency-division multiplexing communication system |
Also Published As
| Publication number | Publication date |
|---|---|
| CN1595924A (en) | 2005-03-16 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| RU2418372C2 (en) | Assessment of orthogonal frequency division multiplex channels | |
| KR100880993B1 (en) | Channel estimation method and apparutus in an ofdm wireless communication system | |
| JP4486992B2 (en) | Wireless communication apparatus and channel estimation and separation method in multi-input system | |
| JP2007089167A (en) | Method of channel estimation in orthogonal frequency division multiplexing system and channel estimator | |
| CN100385824C (en) | Adaptive channel estimation method of MIMO-OFDM system | |
| JP2004159277A (en) | Method and system for detecting symbol of modulated signal received via channel of wireless communications system | |
| Zheng et al. | Channel estimation for frequency-domain equalization of single-carrier broadband wireless communications | |
| CN103152293A (en) | Time and frequency channel estimation in OFDM system | |
| WO2005119953A1 (en) | Method for implementing channel estimate in orthogonal frequency division multiplexing (ofdm) system | |
| CN1437338A (en) | Channel estimation method for orthogonal frequency-division multiplexing communication system | |
| WO2008032849A1 (en) | Wireless communication apparatus | |
| CN101018219B (en) | Space frequency signal processing method | |
| CN102164110A (en) | Method and system for balancing frequency domain | |
| CN102790746B (en) | Channel estimation method for OFDM (orthogonal frequency division multiplexing) system | |
| CN100405790C (en) | Self-adaptive channel estimation method based on least squares criterion of two-dimensional iteration | |
| CN102025662A (en) | Channel estimation method and device for MIMO (multiple input multiple output) OFDM (orthogonal frequency division multiplexing) system | |
| Bhoyar et al. | Leaky least mean square (LLMS) algorithm for channel estimation in BPSK-QPSK-PSK MIMO-OFDM system | |
| Almoneer et al. | Intercarrier interference-aware pilot-aided channel estimation in OFDM systems | |
| JP2003032217A (en) | Ofdm communication device and carrier reproducing method of ofdm signal | |
| CN100588194C (en) | Self-adaptive channel estimation method based on two-dimensional minimum mean square criterion | |
| Bhoyar et al. | Channel estimation for MIMO-OFDM systems | |
| CN111817990B (en) | Channel estimation improvement algorithm based on minimum mean square error in OFDM system | |
| Zhang et al. | Efficient estimation of fast fading OFDM channels | |
| Akhtman et al. | Decision directed channel estimation employing projection approximation subspace tracking | |
| Zhu et al. | A low-complexity channel estimator for OFDM systems in multipath fading channels |
Legal Events
| Date | Code | Title | Description |
|---|---|---|---|
| C06 | Publication | ||
| PB01 | Publication | ||
| C10 | Entry into substantive examination | ||
| SE01 | Entry into force of request for substantive examination | ||
| C14 | Grant of patent or utility model | ||
| GR01 | Patent grant | ||
| C17 | Cessation of patent right | ||
| CF01 | Termination of patent right due to non-payment of annual fee |
Granted publication date: 20080723 |