CN100536448C - MIMO-OFDM channel estimating method for united orthogonal guide frequency design - Google Patents

Abstract

This invention relates to cross conductive frequency MIMO-OFDM signal channel estimation method, which comprises the following steps: designing different emission and receive antenna and different OFDM sub load wave from antenna to keep cross data form and insert the changes of antenna and sub loads into data frame; receive terminal only uses one maximum FFT to extract multi-phase filter to realize sub load wave noisy adding to lower each output end of filter for minimum LS channel parameter estimation.

Description

A kind of MIMO-OFDM channel estimation methods of united orthogonal guide frequency design

Technical field

The present invention relates to a kind of channel estimation methods of wireless communication field, specifically is a kind of channel estimation methods of many reception antenna OFDMs of the many transmit antennas of low complex degree (MIMO-OFDM) system of united orthogonal guide frequency design.

Background technology

Along with the quick growth of personal radio communication demand, wireless technology is experiencing unprecedented opportunity to develop.Any progress of wireless domain research simultaneously all must be tightly round a theme, and that is exactly must solve preferably that bandwidth anxiety, propagation channel that high-speed radiocommunication self faced are abominable, mobile environment is complicated and the quality problems during limited-service.Two-forty forms the wideband frequency selectivity multipath fading of channel, and the continuous variation of scattering object makes channel have selectivity fading effect between multispectral lux-hour around the fast moving of portable terminal or perhaps the communication environments simultaneously.This associating T/F selectivity causes multipath-multispectral the reining in of channel-decline, and can have a strong impact on the overall performance of system.Therefore, the modeling that wireless channel is changed and following the tracks of is estimated and work out that treatment technology is very important in the time of can resisting T/F selectivity decline novel empty concerning mobile radio communications system of future generation.

OFDM (OFDM) is the effective multi-transceiver technology of a kind of frequency, thereby it utilizes the orthogonal narrow-band sub-carriers (bandwidth of each subcarrier is all very narrow, to form the frequency non-selective fading) of some transmits the transmission that low-rate data is realized whole high-speed data concurrently.Many reception antennas of many transmit antennas (MIMO) wireless communication system is under the communication environments that the scattering condition is enriched, can not expand transmission band and realize the higher data transmission efficiency, MIMO can effectively realize diversity in conjunction with Space Time Coding technology or signal processing technology simultaneously, thereby improve the reliability of whole mobile radio system, and for diversity with multiplexingly provide the platform of compromising preferably.These advantage precise channels estimation techniques that make full use of MIMO-OFDM are indispensable.

The technology of estimating about many reception antenna OFDMs of many transmit antennas (MIMO-OFDM) wireless communication system channel roughly can be divided into least square method (LS), linear minimum mean-squared error method (LMMSE), maximum likelihood method (ML) and various blind (total blindness or half-blindness) method of estimation at present.Wherein the computation complexity of least square method is lower, and estimated accuracy is subjected to The noise in the whole transmission system process easily, utilize various channel statistics to come the least-squares estimation value is revised the performance improvement that can obtain to a certain extent, but, can not obtain more real channel statistics information for actual mobile environment.The linear minimum mean-squared error method can obtain reasonable channel parameter estimation precision under the situation of the second-order statistic of channel being done suitable approximate evaluation, but the realization of whole algorithm for estimating need be found the solution an exponent number with subcarrier number and the proportional very big matrix inversion operation of antenna radical, and the complexity that system realizes is very high.Maximum likelihood method utilizes viterbi algorithm to realize the estimation of channel response parameter, and the amount of calculation that need finish rises rapidly along with the increase of whole channel matrix.Blind Channel (or half-blindness) channel estimation methods can improve the band efficiency of whole system to the full extent, do not need (or only need lesser amt) pilot-frequency expense, yet the realization of whole algorithm for estimating requires the actual data volume that transmits very big, and need utilize iterative calculation method, the convergence of algorithm speed limit this algorithm application when real, become wireless mobile environment.

Through the literature search of prior art is found, Chinese patent application number is: 02825159, and name is called: the scattered pilot pattern and the channel estimation methods that are used for the MIMO-OFDM system.This invention provides the number that reduces frequency pilot sign in the MIMO-OFDM communication system and has improved such intrasystem channel estimation methods and equipment.For each transmitting antenna in the OFDM transmitter, frequency pilot sign is encoded, so that be unique for this transmitting antenna.The frequency pilot sign of coding is inserted in the OFDM frame then, and to form network, the network that is used for different transmit antennas uses identical frequency, but departs from single symbol on time domain mutually.At OFDM receiver place, channel response is the symbol for each rhombus center of network, by using two-dimensional interpolation estimative.The estimated channel response is smoothed in frequency domain.The channel response of all the other symbols is estimated by interpolation in frequency domain then.This channel estimation methods uses least square method to carry out initial channel parameter estimation at each pilot tone point, can not effectively overcome or improve the influence of noise to estimated accuracy, and pilot data is not through special design, though for different send-receive antennas concerning position, pilot tone insertion point on its different sub carrier symbol that staggers, can't effectively eliminate different send-receive antennas between the problem of co-channel interference that brings of OFDM frequency sharing.

Summary of the invention

The objective of the invention is at the deficiencies in the prior art, a kind of MIMO-OFDM channel estimation methods of united orthogonal guide frequency design is provided, reach the raising channel estimated accuracy.

The present invention is achieved through the following technical solutions:

A kind of MIMO-OFDM channel estimation methods of united orthogonal guide frequency design, it is characterized in that, design different send-receive antennas between and can both keep the pilot data form of orthogonal property between the right different OFDM subcarriers of same send-receive antenna, simultaneously transmitting terminal with these pilot tones according to the polygon form that changes with the subcarrier number being inserted in the Frame with the send-receive antenna; Only utilize the maximum FFT of a low complex degree to extract the multiphase filter group at receiving terminal and realize windowing effect the noise on the subcarrier, noise variance when carrying out the least square channel parameter estimation to be reduced in each output of bank of filters, thereby raising channel estimated accuracy, for the different send-receive antennas of each output correspondence to and different sub carrier between channel parameter carry out linear interpolation on frequency domain and the time-domain simultaneously, in the time varying channel parameter Estimation that under the situation of priori channel statistics information, can realize preferably under the wireless mobile environment; The specific implementation step is:

(1) channel circumstance that uses according to reality at transmitting terminal, with the completely orthogonal pilot data that designs according to corresponding send-receive antenna to being inserted in the Frame of each transmitting antenna according to parallelogram with sub-carrier positions;

(2) at each reception antenna end send-receive antenna to be estimated is separately extracted pairing pilot data, all pilot datas are sent in the channel estimator, after by bank of filters noise being carried out windowing, corresponding to after to be estimated and the corresponding pilot data output of each subcarrier center frequency point, directly utilize least square method to calculate channel frequency coefficient sample value on the correspondence position;

(3) it is right that the channel coefficients value that again all is estimated is distributed to each send-receive antenna to be estimated according to described insertion position, and each send-receive antenna is exactly the channel frequency coefficient sample value of actual this antenna propagation ducts to resulting channel frequency coefficient sample value;

(4) then with each send-receive antenna between the channel frequency coefficient sample value that obtained respectively in frequency domain and the enterprising line linearity interpolation of time-domain, must all transmit channel coefficients on Data Positions;

(5) at last the channel coefficients value that estimates is given signal processing unit after each root reception antenna, finish the equilibrium of whole system, decoding or closed loop transmission function when empty.

Of the present invention to the effect that design for each send-receive antenna between and all completely orthogonal pilot data between each OFDM subcarrier, realize to each send-receive antenna between the Sampling Estimation function of independently channel frequency response.For the better utilization interpolation smoothly obtain each send-receive antenna between and channel coefficients value on each sub-carrier positions, the pilot tone among the present invention is inserted and is taked the polygon mode to realize.Next utilize respectively and estimate whole wireless traveling time-frequency dual-selection channel coefficient in time-domain linear interpolation and the frequency domain linearity.Each send-receive antenna between pilot tone insert the maximum delay propagation relation in direct ratio of density and channel time-varying rate and channel.Specifically:

Each send-receive antenna to be estimated is to the time shaft pilot tone sampling interval N of interchannel _tSatisfy:

N _t≤1/2f _dT (1)

T is the mark space of OFDM in the formula (1), f _dDoppler frequency expanding value for channel.

Each send-receive antenna to be estimated is to the frequency frequency sampling interval N of interchannel _fSatisfy:

N _f≤1/τ _maxΔF _c (2)

τ in the formula (2) _MaxBe the maximum delay expanding value of channel, Δ F _cBe the frequency interval between each subcarrier.

Being implemented as follows of channel estimating part among the present invention:

At first design a M ₁Rank FIR low pass filter prototype h ₀(n), then to h ₀(n) splitting mutually, processing has

$E_l\left(z^P\right)=\underset{n=0}{\overset{{\mathrm{<figure-callout\; id="1"\; label="M"\; filenames="C20061002695000141.png"\; state="\{\{state\}\}">M</figure-callout>}}_1{/P}^{-1}}{\mathrm{\&Sigma;}}}{h}_o[\mathrm{nP}+l]z^{-\mathrm{lP}}---\left(3\right)$

Thereby obtain corresponding each branch filter coefficient, the pilot data that reception antenna obtains is delivered to the FFT unit with the data after the filtering then through carrying out filtering after the down-sampled processing along separate routes, and counting of FFT equates with the number of OFDM subcarrier.Data after the FFT are carried out least-squares estimation respectively, resulting estimated value is exactly that corresponding send-receive antenna is to the locational channel coefficients value of respective sub, collect good each send-receive antenna between channel domain samples value after directly carry out linear interpolation processing in frequency domain and time-domain, just can obtain corresponding send-receive antenna between whole channel coefficients value.In order to improve the precision of estimation, to carry out threshold judgement to the data after the FFT and handle, just only be considered as valid data, carry out least-squares estimation at last greater than the data of threshold value.

Each send-receive antenna of designing among the present invention between the locational pilot data of respective sub be taken as the complex exponential signal data identical with this subcarrier centre frequency, but different antennae between on frequency, stagger, just:

$X_{\mathrm{pilot}}=\underset{p=0}{\overset{P-1}{\mathrm{\&Sigma;}}}{X}_{\mathrm{pL}}=\underset{p=0}{\overset{P-1}{\mathrm{\&Sigma;}}}{e}^{j\frac{2\mathrm{\&pi;npL}}{N}}---\left(4\right)$

(n，p)∈m _ij，i，j＝1，2

In the formula (4), M _Ij, i, j=1, on behalf of the channel estimating pilot tone between the transmitting antenna i-reception antenna j, 2 insert the T/F shaft position respectively;

After the attenuation through the wireless mobile channel of MIMO, whole pilot data becomes:

$Y_{\mathrm{pilot}}=\underset{p=0}{\overset{P-1}{\mathrm{\&Sigma;}}}\left|H_{\mathrm{ij}}\left(\mathrm{pL}\right)\right|e^{j(\frac{2\mathrm{\&pi;npL}}{N}+S\left(H_{\mathrm{ij}}\left(\mathrm{pL}\right)\right))}+W_{\mathrm{pL}}---\left(5\right)$

(n，p)∈M _ij，i，j＝1，2

W in the formula (5) _PLBe whole H _IjThe zero-mean additive white Gaussian noise that produces in the channel.

The pilot data Y that the reception antenna end is obtained _PilotAfter being sent to the channel estimating part among the present invention, just can directly effectively estimate channel coefficients with least square method, and W _PLThen since the effect of bank of filters can reduce the estimated accuracy influence of least square method.

The present invention has following conspicuous outstanding substantive distinguishing features and remarkable advantage compared with prior art:

The present invention by to different send-receive antennas between and complete orthogonal pilot design between the right different sub carrier position of same antenna, and the reasonable distribution of pilot tone insertion position on whole channel time-frequency grid, utilize a FFT who equates with employing OFDM subcarrier number to realize the least square channel estimation method, and because noise has obtained inhibition preferably, thereby reduced the requirement of least-squares estimation, improved the estimated accuracy of channel parameter noise susceptibility.The channel estimation method that the present invention proposes can reduce greatly because different transmit antennas is used the caused interference of OFDM transmission system of same frequency, can be applicable to the wireless mobile MIMO-OFDM communication system of double selectivity slow fading.

Description of drawings

Fig. 1 is for using the system works flow process figure of the inventive method.

Fig. 2 is the pilot tone insertion position and the inserted mode schematic diagram of the inventive method.

Figure (a) is the pilot tone insertion position and the inserted mode schematic diagram of transmitting antenna 1 among Fig. 2.

Figure (b) is the pilot tone insertion position and the inserted mode schematic diagram of transmitting antenna 2 among Fig. 2.

Fig. 3 is the channel estimating schematic diagram in the inventive method.

Embodiment

A preferred embodiment of the present invention is: referring to Fig. 1, the MIMO-OFDM channel estimation methods step of this united orthogonal guide frequency design is: design different send-receive antennas between and can both keep the pilot data form of orthogonal property between the right different OFDM subcarriers of same send-receive antenna, simultaneously transmitting terminal with these pilot tones according to the polygon form that changes with the subcarrier number being inserted in the Frame with the send-receive antenna; Only utilize the maximum FFT of a low complex degree to extract the multiphase filter group at receiving terminal and realize windowing effect the noise on the subcarrier, noise variance when carrying out the least square channel parameter estimation to be reduced in each output of bank of filters, thereby raising channel estimated accuracy, for the different send-receive antennas of each output correspondence to and different sub carrier between channel parameter carry out linear interpolation on frequency domain and the time-domain simultaneously, in the time varying channel parameter Estimation that under the situation of priori channel statistics information, can realize preferably under the wireless mobile environment; The specific implementation step is:

(1) channel circumstance that uses according to reality at transmitting terminal, with the completely orthogonal pilot data that designs according to corresponding send-receive antenna to being inserted in the Frame of each transmitting antenna according to parallelogram with sub-carrier positions;

(2) at each reception antenna end send-receive antenna to be estimated is separately extracted pairing pilot data, all pilot datas are sent in the channel estimator, after by bank of filters noise being carried out windowing, corresponding to after to be estimated and the corresponding pilot data output of each subcarrier center frequency point, directly utilize least square method to calculate channel frequency coefficient sample value on the correspondence position;

(3) it is right that the channel coefficients value that again all is estimated is distributed to each send-receive antenna to be estimated according to described insertion position, and each send-receive antenna is exactly the channel frequency coefficient sample value of actual this antenna propagation ducts to resulting channel frequency coefficient sample value;

(4) then with each send-receive antenna between the channel frequency coefficient sample value that obtained respectively in frequency domain and the enterprising line linearity interpolation of time-domain, transmit channel coefficients on Data Positions thereby obtain all;

(5) at last the channel coefficients value that estimates is given signal processing unit after each root reception antenna, finish the equilibrium of whole system, decoding or closed loop transmission function when empty.

In the above-mentioned step (1), the number that the time shaft in every frame data inserts pilot number is the time interval N that pilot tone is inserted _t, with the rate of change of channel relation in direct ratio, be specially:

Nt≤1/2f _dT

T is the OFDM mark space in the following formula, f _dBe the doppler frequency expanding value.

In the above-mentioned step (1), the number of the frequency axis insertion pilot data in every frame data is promptly inserted the frequency interval N of pilot tone _f, the inversely proportional relation of maximum delay amount with channel is specially:

N _f≤1/τ _maxΔF _c

τ in the following formula _MaxBe the maximum multipath retardation of channel, Δ F _cBe the frequency interval between the subcarrier.

Each send-receive antenna of designing between the locational pilot data of respective sub be taken as the complex exponential signal data identical with this subcarrier centre frequency, but different antennae between on frequency, stagger, just:

$X_{\mathrm{pilot}}=\underset{p=0}{\overset{P-1}{\mathrm{\&Sigma;}}}{X}_{\mathrm{pL}}=\underset{p=0}{\overset{P-1}{\mathrm{\&Sigma;}}}{e}^{j\frac{2\mathrm{\&pi;npL}}{N}}$

(n，p)∈m _ij，i，j＝1，2

In the following formula, M _Ij, i, j=1, on behalf of the channel estimating pilot tone between the transmitting antenna i-reception antenna j, 2 insert the T/F shaft position respectively;

The effect of the frequency selective fading of process mobile radio channel, then pilot data becomes:

$Y_{\mathrm{pilot}}=\underset{p=0}{\overset{P-1}{\mathrm{\&Sigma;}}}\left|H_{\mathrm{ij}}\left(\mathrm{pL}\right)\right|e^{j(\frac{2\mathrm{\&pi;npL}}{N}+S\left(H_{\mathrm{ij}}\left(\mathrm{pL}\right)\right))}+W_{\mathrm{pL}}$

(n，p)∈M _ij，i，j＝1，2

In the formula, W _PLBe whole H _IjThe zero-mean additive white Gaussian noise that produces in the channel.

Above-mentioned step (2), channel estimating part specific implementation is: the FIR low pass filter prototype h that at first designs M1 rank ₀(n), then to h ₀(n) carrying out heterogeneous decomposition has:

$E_l\left(z^P\right)=\underset{n=0}{\overset{{\mathrm{<figure-callout\; id="1"\; label="M"\; filenames="C20061002695000141.png"\; state="\{\{state\}\}">M</figure-callout>}}_1/P^{-1}}{\mathrm{\&Sigma;}}}{h}_o[\mathrm{nP}+l]z^{-\mathrm{lP}}$

Thereby obtain corresponding each branch filter coefficient, the pilot data that reception antenna obtains is delivered to the FFT unit with the data after the filtering then through carrying out filtering after the down-sampled processing along separate routes, and counting of FFT equates with the number of OFDM subcarrier; Data after the FFT are carried out least-squares estimation respectively, resulting estimated value is exactly that corresponding send-receive antenna is to the locational channel coefficients value of respective sub, collect good each send-receive antenna between channel domain samples value after directly carry out linear interpolation processing in frequency domain and time-domain, just can obtain corresponding send-receive antenna between whole channel coefficients value; In order to improve the precision of estimation, to carry out threshold judgement to the data after the FFT and handle, just only be considered as valid data, carry out least-squares estimation at last greater than the data of threshold value.

As shown in Figure 1, the data that transmit are at first handled through serial, join together to form Frame at the pilot data that carries out on Space Time Coding and the correspondence position then.After each Frame is handled through IFFT respectively again, become the signal that radiofrequency signal launches on each antenna of receiving terminal from two antennas and at first will extract through the pilot signal of fading channel effect separately, the pilot signal that extracts is sent into channel estimating unit among the present invention according to the pilot tone insertion position.Through obtain after channel coefficients estimation and time-domain and the frequency domain linear interpolation each send-receive antenna between channel coefficient vector, utilize the channel vector that estimates to realize frequency domain equalization and decoding when empty then.

As shown in Figure 2, the ripple FIR filter order that waits of channel estimating part is taken as 128, under to adopt factor Q be 4, N is 64, F gets 2, directly carry out least square (LS) through each data after the FFT processing and estimate that the data that estimate are done the Channel Transmission matrix that linear interpolation processing just can obtain whole M IMO system after distributing according to the pilot tone insertion position.

As shown in Figure 3, if whole transmission system adopts two transmit antennas-two reception antenna, the characterisitic parameter of individual channel is consistent, that is to say that four channel vectors of needs estimation take the mode of rhombus insertion pilot tone to estimate and smoothing processing respectively.The size of rhombus will be adjusted according to the channel condition of reality.

Claims (5)

1, a kind of MIMO-OFDM channel estimation methods of united orthogonal guide frequency design, it is characterized in that, design different send-receive antennas between and can both keep the pilot data form of orthogonal property between the right different OFDM subcarriers of same send-receive antenna, simultaneously transmitting terminal with these pilot tones according to the polygon form that changes with the subcarrier number being inserted in the Frame with the send-receive antenna; Only utilize the maximum FFT of a low complex degree to extract the multiphase filter group at receiving terminal and realize windowing effect the noise on the subcarrier, noise variance when carrying out the least square channel parameter estimation to be reduced in each output of bank of filters, thereby raising channel estimated accuracy, for the different send-receive antennas of each output correspondence to and different sub carrier between channel parameter carry out linear interpolation on frequency domain and the time-domain simultaneously, in the time varying channel parameter Estimation that under the situation of priori channel statistics information, can realize preferably under the wireless mobile environment; The specific implementation step is:

(1) channel circumstance that uses according to reality at transmitting terminal, with the completely orthogonal pilot data that designs according to corresponding send-receive antenna to being inserted in the Frame of each transmitting antenna according to parallelogram with sub-carrier positions;

(2) at each reception antenna end send-receive antenna to be estimated is separately extracted pairing pilot data, all pilot datas are sent in the channel estimator, after by bank of filters noise being carried out windowing, corresponding to after to be estimated and the corresponding pilot data output of each subcarrier center frequency point, directly utilize least square method to calculate channel frequency coefficient sample value on the correspondence position;

(3) it is right that the channel coefficients value that again all is estimated is distributed to each send-receive antenna to be estimated according to described insertion position, and each send-receive antenna is exactly the channel frequency coefficient sample value of actual this antenna propagation ducts to resulting channel frequency coefficient sample value;

(4) then with each send-receive antenna between the channel frequency coefficient sample value that obtained respectively in frequency domain and the enterprising line linearity interpolation of time-domain, transmit channel coefficients on Data Positions thereby obtain all;

(5) at last the channel coefficients value that estimates is given signal processing unit after each root reception antenna, finish the equilibrium of whole system, decoding or closed loop transmission function when empty.

2, the MIMO-OFDM channel estimation methods of united orthogonal guide frequency design according to claim 1 is characterized in that, in the described step (1), the number that the time shaft in every frame data inserts pilot number is the time interval N that pilot tone is inserted _t, with the rate of change of channel relation in direct ratio, be specially:

N _t≤1/2f _dT

T is the OFDM mark space in the following formula, f _dBe the doppler frequency expanding value.

3, the MIMO-OFDM channel estimation methods of united orthogonal guide frequency design according to claim 1 is characterized in that, in the described step (1), the number of the frequency axis insertion pilot data in every frame data is promptly inserted the frequency interval N of pilot tone _f, the inversely proportional relation of maximum delay amount with channel is specially:

N _f≤1/τ _maxΔF _c

τ in the following formula _MaxBe the maximum multipath retardation of channel, Δ F _cBe the frequency interval between the subcarrier.

4, the MIMO-OFDM channel estimation methods of united orthogonal guide frequency design according to claim 1, it is characterized in that, each send-receive antenna of designing between the locational pilot data of respective sub be taken as the complex exponential signal data identical with this subcarrier centre frequency, but different antennae between on frequency, stagger, just

$X_{\mathrm{pilot}}=\underset{p=0}{\overset{P-1}{\mathrm{\&Sigma;}}}{X}_{\mathrm{pL}}=\underset{p=0}{\overset{P-1}{\mathrm{\&Sigma;}}}{e}^{j\frac{2\mathrm{\&pi;npL}}{N}}$

(n，p)∈M _ij，i，j＝1，2

In the following formula, M _Ij, i, j=1, on behalf of the channel estimating pilot tone between the transmitting antenna i-reception antenna j, 2 insert the T/F shaft position respectively;

The effect of the frequency selective fading of process mobile radio channel, then pilot data becomes:

(n，p)∈M _ij，i，j＝1，2

In the formula, W _PLBe whole H _IjThe zero-mean additive white Gaussian noise that produces in the channel.

5, the MIMO-OFDM channel estimation methods of united orthogonal guide frequency design according to claim 1 is characterized in that, described step (2), and channel estimating part specific implementation is: the FIR low pass filter prototype h that at first designs M1 rank ₀(n), then to h ₀(n) carrying out heterogeneous decomposition has

$E_l\left(z^P\right)=\underset{n=0}{\overset{M_1/P-1}{\mathrm{\&Sigma;}}}{h}_o[\mathrm{nP}+l]z^{-\mathrm{lP}}$

Thereby obtain corresponding each branch filter coefficient, the pilot data that reception antenna obtains is delivered to the FFT unit with the data after the filtering then through carrying out filtering after the down-sampled processing along separate routes, and counting of FFT equates with the number of OFDM subcarrier; Data after the FFT are carried out least-squares estimation respectively, resulting estimated value is exactly that corresponding send-receive antenna is to the locational channel coefficients value of respective sub, collect good each send-receive antenna between channel domain samples value after directly carry out linear interpolation processing in frequency domain and time-domain, just can obtain corresponding send-receive antenna between whole channel coefficients value; In order to improve the precision of estimation, to carry out threshold judgement to the data after the FFT and handle, just only be considered as valid data, carry out least-squares estimation at last greater than the data of threshold value.

Images