CN100385824C - Adaptive channel estimation method of MIMO-OFDM system - Google Patents

Abstract

The present invention discloses a channel estimation method for MIMO-OFDM systems, which comprises the transmission of data frame structures and channel estimation. Transmitted data is divided into training OFDM symbols and data transmission OFDM symbols, wherein the training symbols are composed of training pilot symbols and 0 symbols and are used for establishing the initial parameter of channel estimation, and the data transmission OFDM symbols are composed of reference pilot symbols and data symbols; an adaptive algorithm is adopted for the channel estimation. The method of the present invention has the advantage of high spectrum efficiency, and an algorithm of a receiver has the advantage of low calculating complexity.

Description

The adaptive channel estimation method of a kind of MIMO-OFDM system

Technical field:

The invention belongs to wireless communication field, it is particularly related to the channel estimating in the MIMO-OFDM system.

Background technology:

Future broadband wireless communication systems requires to have higher spectrum efficiency, so that data transport service reliably to be provided at a high speed.Multiple-input, multiple-output (MIMO) wireless communication system can significantly improve capacity of communication system or communication reliability, and the compromise of the two also can be provided.(see document [L.Zheng, et al, " Diversity and Multiplexing:A Fundamental Tradeoff inMultiple-Antenna Channels; IEEE Trans.On Inform.Theory, vol.49, No.5; pp.1073-1096, May2003]).Simultaneously, future mobile communication system should be broadband system.The typical problem of broadband system is a multipath effect, for the simplified receiver design, the OFDM modulation is considered to a kind of competitive solution, and OFDM is used for mimo system is the advantage that the MIMO-OFDM system can obtain MIMO and OFDM technology simultaneously, thereby is considered to have wide practical use in following wireless communications application.(see document [D.Agrawal, et al., " Space-time coded OFDM for highdata-rate wireless communication over wideband channels; " in IEEE Vehi.Tech.Conference, vol.3, pp.2232--2236,1998.] and [Y. (G.) Li, et.al., " Transmit diversity for OFDM systems and itsimpact on high-rate data wireless networks, " IEEE J.Select.Areas Commun., vol.17, pp.1233-1243, July 1999.]

The frequency-region signal model of MIMO-OFDM system is (as Fig. 1, shown in Figure 2):

$Y_i[n,k]=\underset{q=1}{\overset{M_T}{\mathrm{\Σ}}}{H}_{i,q}[n,k]S_q[n,k]+V_i[n,k]$ i＝1，2，…，M _R，k＝0，1，…，N-1 (1)

S wherein _q[n, k] is the q transmitting antenna signal that the k subcarrier sends in n OFDM symbol; N is a positive integer, expression OFDM sub-carrier number; Make h _{I, q}[n, l] is the path fading coefficient of l bar multipath component between interior q transmitting antenna to the i reception antenna of n OFDM symbol: $H_{i,q}[n,k]=\underset{l=0}{\overset{L-1}{\mathrm{\&Sigma;}}}{h}_{i,q}[n,l]e^{-\mathrm{<figure-callout\; id="2"\; label="j"\; filenames="C20041002186400182.png"\; state="\{\{state\}\}">j</figure-callout>}\frac{2\mathrm{\&pi;kl}}{N}}$ For channel frequency response wherein L be that positive integer is represented channel exponent number; V _i[n, k] is zero-mean additive white Gaussian noise, satisfies

$E\left\{V_{i\&prime;}\right[n^{\&prime;},k^{\&prime;}\left]V_i^{*}\right[n,k\left]\right\}={\mathrm{\&sigma;}}_{\mathrm{<figure-callout\; id="2"\; label="V"\; filenames="C20041002186400182.png"\; state="\{\{state\}\}">V</figure-callout>}}^2\mathrm{\&delta;}[i^{\&prime;}-i]\mathrm{\&delta;}[n^{\&prime;}-n]\mathrm{\&delta;}[m^{\&prime;}-m]---\left(2\right)$

Wherein δ [m-n] is that 0 o'clock value is 1 at m-n, and other situations are 0.σ _V ²Expression noise variance.

In various MIMO-OFDM system, the estimation of wideband MIMO temporal frequency double selectivity fading channel and tracking are prerequisite and the keys that realizes the optimum decoding of receiver.The channel estimation technique of traditional single-shot list receipts (SISO) ofdm system has been obtained great successes and (has been seen Y. (G.) Li, " Pilot-Symbol-Aided Channel Estimation for OFDM inWireless Systems ", IEEE Trans On Vehicular Technology vol..49, no.4, pp1207-1215, July 2000 and list of references thereof).Recently, the channel estimating of MIMO-OFDM system also causes widely to be paid close attention to, and the main MIMO-OFDM system channel estimation method that proposes mainly contains at present:

[1] channel estimation scheme that is applicable to decision-feedback or optimal training sequence under the MMSE criterion (is seen document Y. (G.) Li, " Simplified Channel Estimation for OFDM Systems with Multiple Transmit Antennas ", IEEETrans.On Wireless Comm., vol.1, pp.67-75, Jan.2002)

[2] based on frequency pilot sign the time domain interpolating method (seeing document K.Lee and D.B.Williamos; " Pilot-Symbol-Assisted Channel Estimation for Space-Time Coded OFDMSystems; " EURASIP Journal on Applied Signal Processing 2002:5,507-516)

[3] (see document H. based on the blind Channel Estimation algorithm of methods such as subspace

, et al., " Blind ChannelIdentification and Equalization in OFDM-Based Multiantenna Systems ", IEEE Trans On SignalProcessing, vol.50, pp96-109, Jan 2002)

Said method is applicable to different occasions respectively, but also some significant disadvantage of each tool.Method [1] has the performance of optimization but its amount of calculation is big, require the receiver disposal ability strong, with high costs, be difficult to practical to practical wireless communication systems, wherein the method based on training sequence requires frequently to send training sequence in the rapid fading application scenario, causes loss of spectral efficiency very big.The typical problem of method [2] is that its loss of spectral efficiency is bigger under the rapid fading occasion, and its loss of spectral efficiency is directly proportional with number of transmit antennas owing to the restriction of Nyquist sampling rate to pilot symbol interval.The typical problem of method [3] is that computational complexity is very big, and can have problems such as phase ambiguity, and practical value is not high.

Summary of the invention

The purpose of this invention is to provide a kind of MIMO-OFDM system channel estimation approach, it has higher spectrum efficiency, and its receiver algorithm has lower computational complexity.

For convenience of explanation, the following term of specific justice:

Training frequency guide symbol: the symbol that transmitting-receiving two-end is all known, specific subcarrier by first OFDM symbol (being training OFDM symbol) of Frame transmits, and it act as and makes receiver channel estimation module estimated service life least square (LS) method estimate the channel initial value.

Reference pilot symbols: the symbol that transmitting-receiving two-end is all known, the specific subcarrier transmission by the OFDM symbol (being data transmission OFDM symbol) behind first OFDM symbol of Frame is used to produce the required input signal vector of adaptive channel tracing algorithm.

Main multipath component is selected (STC): estimate (estimated value by more multipath fading coefficient is formed) from the bigger channel time domain impulse response of exponent number, according to the principle that reduces evaluated error, select the estimated result of wherein less multipath fading coefficient as channel time domain impulse response, its purpose is to reduce the multipath fading coefficient, reduces number of parameters and reduces evaluated error.Its implementation is seen Y. (G.) Li, Simplified Channel Estimation for OFDM Systems With MultipleTransmit Antennas, IEEE Trans.On Wirel.Comm.Vol.1, pp.67-75, Jan.2002 and list of references thereof.

Channel impulse response brachymemma: estimate (estimated value by more multipath fading coefficient is formed) from the bigger channel time domain impulse response of exponent number, according to the principle that shortens channel impulse response length, from the shortest multipath fading coefficient of time delay, extract of the estimation of the multipath fading coefficient of the short given number of time delay as channel impulse response.

Least mean square algorithm (LMS algorithm): a kind of typical adaptive algorithm, but its realization reference literature Simon Haykin Adaptive Filter Theory, Third edition, Prentice Hall1998 based on minimum mean square error criterion (MMSE).

Iterative least square algorithm (RLS algorithm): a kind of typical adaptive algorithm, but its realization reference literature Simon Haykin Adaptive Filter Theory, Third edition, Prentice Hall 1998 based on criterion of least squares (LS).

QR-RLS algorithm: a kind of implementation method of RLS algorithm, but its realization reference literature Simon Haykin AdaptiveFilter Theory, Third edition, Prentice Hall 1998.

A kind of new self-adaption MIMO-OFDM channel estimation methods of the present invention is characterized in that it comprises make a start step and receiving end step:

The described step of making a start is carried out according to step 1 and step 2:

Step 1: introduce training frequency guide symbol and reference pilot symbols to form data frame structure

To any transmitting antenna, transmit data according to a minute frame mode, every frame comprises N _t(N _tBe positive integer) individual OFDM symbol; First OFDM symbol is training OFDM symbol (n=0, n represent OFDM symbol sequence number) in the frame; N after the training OFDM symbol in the frame _t-1 OFDM symbol be data transmission OFDM symbol (n=1,2 ..., N _t-1).(as shown in Figure 3); Insert training frequency guide symbol and insert reference pilot symbols according to following method:

Insert training frequency guide symbol: in training OFDM symbol, for q (q=1 wherein, 2 ..., M _T, M _TThe expression number of transmit antennas is positive integer) transmitting antenna, its first transmit training frequency guide symbol subcarrier number be q-1, thereafter every D _F0Individual subcarrier inserts a training frequency guide symbol; To this transmitting antenna, at non-training frequency guide symbol subcarrier place, the transmission amplitude is 00 symbol; (as shown in Figure 4).

Insert reference pilot symbols: in any data transmission OFDM symbol, all transmitting antennas use identical subcarrier to transmit reference pilot symbols; To any transmitting antenna, the subcarrier sequence number that transmits the 1st reference pilot symbols is k ₀(0≤k ₀＜D _F1, D _F1Be positive integer), thereafter every D _F1Individual subcarrier inserts a reference pilot symbols; To any transmitting antenna,, transmit the data symbol of Space Time Coding output at non-reference pilot symbols subcarrier place.The subcarrier sequence number that transmits reference pilot symbols is k _p, p=0,1 ..., P-1,

N is the OFDM sub-carrier number;

Step 2: the frame structure that forms according to step 1 sends.

Described receiving end step was carried out according to following stage 1 and stage 2:

Stage 1: estimate channel at the training OFDM symbol place of every frame, be used for the initial parameter that adaptive channel is followed the trail of during the transfer of data in this frame, carry out in turn according to A, B and three steps of C to obtain:

Steps A: obtain the estimated value of the frequency response of all training frequency guide symbol subcarriers of current channel with least square (being LS) estimation:

Step B: to the result of steps A, use frequency domain-＞spatial transform estimates that all transmitting antennas to channel impulse response between all reception antennas (remembers in its channel impulse response being estimated as of l bar multipath component multipath fading coefficient between q transmitting antenna to the i reception antenna in n OFDM symbol

0≤l≤N-1) wherein.I=1 wherein, 2 ..., M _R, M _RBe the reception antenna number.

Step C: obtain channel master multipath component time of delay and extract main multipath component fading coefficients:

Utilize result of calculation among the step B, use direct truncation method of channel impulse response or main multipath component to select (STC) method, to the channel between the extremely any reception antenna i of any transmitting antenna q, from its channel multi-path fading coefficients

The middle selection (

Be positive integer) bar master multipath, store its time of delay (being unit with the sampling period) and be l _{I, q, k}Fading coefficients with respective paths

Wherein $k=\mathrm{1,2},\&CenterDot;\&CenterDot;\&CenterDot;,\stackrel{\&LeftRightArrow;}{L}.$

Stage 2: adaptive channel is followed the trail of during the transfer of data:

To the n in the frame (n=1,2 ..., N _t-1) individual data transmission OFDM symbol, to arbitrary reception antenna i wherein, use following channel estimation methods (all reception antennas are repeated following steps A, B and C):

Steps A: adaptive channel track algorithm initial value is set:

Adaptive channel is set estimates that initial value is the final value that n-1 OFDM symbol estimated:

n＝1，2，…，N _t-1 (3)

Wherein:

Represent in n the OFDM symbol that through after p iteration, all transmitting antennas are to the estimated result of channel between the reception antenna i,

Represent in n the OFDM symbol that through after P iteration, all transmitting antennas are to the final result of channel estimating between the reception antenna i,

Represent in n the OFDM symbol that through after P iteration, q transmitting antenna is to the final result of channel estimating between the reception antenna i, wherein

Step B: adaptive channel is followed the tracks of:

Arbitrarily the reference pilot symbols subcarrier is all corresponding to an iteration of adaptive algorithm, to p (0≤p≤P-1, P are the reference pilot symbols number) arbitrarily:

The reference signal of adaptive algorithm is Y _i[n, k _p] (received signal at p reference pilot subcarrier of i reception antenna place during n OFDM symbol), wherein k _pFor transmitting the subcarrier sequence number of p reference pilot symbols;

The input signal vector w of adaptive algorithm _p[n] is:

$w_p\left[n\right]={\left[{\mathrm{<figure-callout\; id="1"\; label="w"\; filenames="C20041002186400171.png,C20041002186400182.png"\; state="\{\{state\}\}">w</figure-callout>}}_{1,p}^T\right[n],{\mathrm{<figure-callout\; id="2"\; label="w"\; filenames="C20041002186400182.png"\; state="\{\{state\}\}">w</figure-callout>}}_{2,p}^T[n],\&CenterDot;\&CenterDot;\&CenterDot;,w_{M_T,p}^T[n\left]\right]}^T---\left(6\right)$

Wherein

The value of representing p reference pilot symbols of q transmitting antenna in n the data transmission OFDM, receiving terminal is known should to be worth;

Adaptive channel is estimated: can adopt typical adaptive algorithm (comprise the LMS algorithm, in RLS algorithm and the QR-RLS algorithm any) to adjust

In the current data transmission OFDM symbol, iteration is carried out (P is the number of reference pilot symbols in each OFDM symbol) altogether P time, obtains

As the time domain channel results estimated in this OFDM symbol, promptly

Step C: utilize step B adaptive channel results estimated to estimate the order of all channel sub-carrier frequencies responses, split according to the combination of (4) and (5) formula

Obtain of the estimation of all transmitting antennas to each the main multipath component of channel between reception antenna i.The estimated value of remembering main multipath component between q transmitting antenna to the i reception antenna is

Use and to obtain during n the OFDM symbol channel k sub-carrier frequency domain channel estimating between q transmitting antenna to the i reception antenna behind the Fourier transform (perhaps FFT).

Need to prove that the present invention program's realization is to use on the insertion of frequency pilot sign and the receiving terminal Processing Algorithm on the basis of method proposed by the invention, adopts the typical case to realize based on the method for the MIMO-OFDM system of frequency pilot sign assisted channel estimation.

The present invention designs basic principle:

1) for different transmit antennas, adopt different sub carrier to transmit the training pilot signal at the training OFDM symbol place, mimo channel can be converted to a plurality of independent and mutually orthogonal SISO channels, thus all antennas of estimating the training symbol place by means of simple LS method to channel response;

2) therefore different decline and the noise jamming of all multipath component experience of multipath channel reject the accuracy that the low multipath component of signal to noise ratio can effectively improve channel estimating.

3) traditional channel estimator based on training only is suitable for the low occasion of channel time-varying speed, and can effectively follow the tracks of time varying channel by means of the channel estimation methods that inserts ground training frequency guide symbol and reference pilot symbols that distributes.

4) the preferred channels estimator often needs channel statistic property as prior information, and this often needs the time overhead of training in actual applications, and adaptive channel is estimated to avoid this requirement.Adaptive channel has estimated also to avoid the mis-behave problem of preferred channels estimation when the channel statistical characteristic variations in addition.

5) the adaptive algorithm reference signal that needs some could restrain, channel response is being a continually varying between the OFDM symbol continuously simultaneously, thereby the initial value that the channel estimation results that will go up an OFDM symbol is estimated as current OFDM symbol self adaptation can reduce required reference data number, thereby can avoid the deterioration of spectrum efficiency.

Innovation part of the present invention is:

1) send data and be made up of 2 parts, channel estimating also is divided into 2 stages: send data and be divided into training OFDM symbol and data transmission OFDM symbol (as shown in Figure 3).Wherein training OFDM symbol is made up of training frequency guide symbol and 0 symbol, is used to set up the initial parameter of channel estimating; Data transmission OFDM symbol is made up of reference pilot symbols and data symbol.The difference of reference pilot symbols and traditional frequency pilot sign is that its effect is the input signal that is used to produce the adaptive channel algorithm for estimating among the present invention, the broad-band channel that becomes during with self-adoptive trace (note: but not be used for combining to estimate channel frequency response with interpolation method).About the significant advantage of the another one of reference pilot symbols be, the reference pilot symbols of different transmit antennas occupies identical subcarrier, thereby frequency expense and number of transmit antennas that reference pilot symbols is caused are irrelevant, and does not need particular design.The insertion pattern of its training frequency guide symbol and reference pilot symbols as shown in Figure 4.

2) channel estimation method adopts adaptive algorithm: adaptive algorithm have can be along with channel variation the function of dynamic tracking channel parameter, have characteristics such as the channel statistical of need not characteristic, amount of calculation are little.

The present invention has the following advantages:

1) channel estimating is used the initial value of training OFDM symbol as channel estimating, accelerates the convergence rate of adaptive channel algorithm for estimating;

2) all transmitting antennas use same sub-carrier to transmit reference pilot symbols during the data transmission OFDM symbol; And its effect is to be used to produce the input signal of adaptive algorithm and reference signal but not as interpolation, to have high spectrum utilization;

3) adopt adaptive algorithm to follow the tracks of channel, need not the channel statistical characteristic;

4) channel tracking carries out in time domain;

5) the inventive method can adopt the error of main multipath component selection technology with reduction channel estimating initial value, and can reduce the number of the channel parameter that needs tracking, thereby reduces operand and improve accuracy of estimation.

Description of drawings

Fig. 1 is a typical MIMO-OFDM system emission block diagram

S wherein _q[n, k] is the signal that q transmitting antenna k subcarrier transmits during the nOFDM symbol;

It is q transmitting antenna during n the OFDM symbol

The time-domain signal that sampling instant transmits; N represents that (n is an integer, 0≤n＜N for the sequence number of OFDM symbol in the frame _T), k represents OFDM subcarrier sequence number (k is an integer, and 0≤k＜N, N represent the OFDM total number of sub-carriers),

The expression sampling instant (

Be integer, $-N_g\&le;\tilde{m}<N,$ N wherein _gExpression OFDM cyclic prefix samples is counted).Frame wherein forms module and finishes the function of inserting training symbol and frequency pilot sign, and among the present invention, frame forms module need insert special training frequency guide symbol and reference pilot symbols.

Fig. 2 is typical MIMO-OFDM receiver block diagram

Wherein

Represent i root reception antenna during n the OFDM symbol

The sampled signal that constantly receives, Y _i[n, k] separates the decoder input signal of decoding when being used for sky for i reception antenna k subcarrier during n OFDM symbol of receiver; Wherein i is a positive integer, 1≤i≤M _R, M _RBe the reception antenna number; Channel estimation module wherein uses the frequency pilot sign that extracts from received signal to estimate the mimo channel frequency response.

Fig. 3 is a data transmission frame structure of the present invention

Total N in each frame _tIndividual OFDM symbol, wherein first (n=0) OFDM symbol is a training OFDM symbol, is total to N thereafter _t-1 OFDM symbol be data transmission OFDM symbol (n=1,2 ..., N _t-1).

Fig. 4 is training frequency guide symbol, the reference pilot symbols pattern of this programme design

Provided among the figure transmitting antenna q (q=1 wherein, 2 ..., M _T, M _TBe number of transmit antennas) training frequency guide symbol and the formation pattern of reference pilot symbols; Training OFDM symbol, all the other common N are shown in first tabulation among the figure _tData transmission OFDM symbol is shown in-1 tabulation; N represents the sequence number of OFDM symbol, 0≤n≤N _t-1), k represents the subcarrier sequence number, 0≤k≤N-1, N _tBe every frame OFDM symbol numbers; D _F0In training OFDM symbol, any transmitting antenna two adjacent training frequency guide symbol subcarrier spacings; D _F1Subcarrier spacing for two adjacent reference pilot symbols in any transmitting antenna data transmission OFDM symbol;

The expression training frequency guide symbol,

The expression reference pilot symbols,

The expression data symbol,

Represent 0 symbol.

Fig. 5 is a channel self-adapting tracing scheme schematic diagram of the present invention

According to current OFDM Symbol Type, when being training OFDM symbol, use least square (LS) algorithm to estimate current channel response and select main multipath component, first stage that this receiving terminal is handled for first OFDM symbol in the frame; When current OFDM symbol is data transmission OFDM symbol, enter receiving terminal channel estimating second stage, promptly use the adaptive channel track algorithm to follow the tracks of channel.

Fig. 6 is to use the systematic function emulation of the present invention program's channel estimating to give an example

Provided bit error rate (BER) the computer policy performance curve that uses the present invention program's a special case among the figure, wherein SNR represents signal to noise ratio, and esti represents to use the inventive method results estimated, and ideal represents the ideal communication channel results estimated.What adaptive algorithm adopted is the LMS algorithm.Provided the system's performance of BER under parameter is selected in the following embodiment among the figure.Dotted line wherein is the ideal communication channel estimated performance, and solid line is the performance of this paper scheme.

Embodiment

Below to provide a concrete MIMO-OFDM configuration down, the performing step of this patent.Parameter in this example does not influence generality of the present invention.

The treatment step of making a start:

Steps A: introduce training frequency guide symbol and reference pilot symbols to form data frame structure

To each transmit antennas, transfer of data transmits according to a minute frame mode; First OFDM symbol of every frame is a training OFDM symbol; It is data transmission OFDM symbol thereafter; The OFDM symbol numbers that each frame comprises is N _t=40.

Training frequency guide symbol is arranged in training OFDM symbol, and different transmit antennas adopts different sub carrier to transmit training frequency guide symbol.(shown in figure (4)).The subcarrier that its q transmitting antenna p training frequency guide symbol is distributed is

${\stackrel{\&OverBar;}{k}}_{q,p}=q-1+\mathrm{<figure-callout\; id="0"\; label="p\; D\; f"\; filenames="C20041002186400171.png"\; state="\{\{state\}\}">p</figure-callout>}{D}_f{0}_{},$ $p=\mathrm{0,1},\&CenterDot;\&CenterDot;\&CenterDot;,\stackrel{\&OverBar;}{P}-1,$ q＝1，2，…，M _T

Wherein

Training frequency guide symbol number for the single transmission antennas transmit in training OFDM symbol place;

Wherein q is the transmitting antenna sequence number, 1≤q≤M _T, select number of transmit antennas M in this example _T=2, OFDM modulated carrier number is N=1024; Non-training frequency guide symbol place's transmission amplitude is 00 symbol in the training OFDM symbol,

Transmit in the OFDM symbol at arbitrary data, all transmitting antennas all adopt identical subcarrier to transmit reference pilot symbols.The sequence number that is used to transmit the subcarrier of reference pilot symbols is:

k _p＝pD _f1+k ₀，p＝0，1，…，P-1

Wherein total reference subcarrier number is

Wherein select D in this example _F1=16, k ₀=0.Transmit the coding result of space-time grid coding at other subcarrier place.

Step B: the frame structure according to steps A sends;

Channel estimating is carried out at receiving terminal; Carry out according to stage 1 and 2 two stages of stage:

Stage 1: estimate channel at the training OFDM symbol place of every frame, be used for the initial parameter that adaptive channel is followed the trail of during the transfer of data in this frame, carry out in turn according to A, B and three steps of C to obtain:

Steps A: obtain the estimated value of the frequency response of all training frequency guide symbol subcarriers of current channel with least square (being LS) estimation:

i＝1，2，…，M _R，q＝1，2，…，M _T， $p=\mathrm{0,1},\&CenterDot;\&CenterDot;\&CenterDot;,\stackrel{\&OverBar;}{P}-1,$ n＝0

Wherein

For transmitting antenna q to reception antenna i place The estimation of the frequency response at subcarrier place,

It is i reception antenna the in n the OFDM symbol

The subcarrier received signal;

Be that the q transmitting antenna is

The known training signal that individual subcarrier transmits; Wherein i is the reception antenna sequence number, 1≤i≤M _R, select M in this example _R=2.

Step B: use frequency domain-＞spatial transform estimates that all transmitting antennas are to channel impulse response between all reception antennas:

i＝1，2，…，M _R，q＝1，2，…，M _T，n＝0

Wherein:

Be n OFDM symbol place in the frame, the estimation of channel impulse response between q transmitting antenna to the i reception antenna,

Be n OFDM symbol place in the corresponding frame, the estimation of channel l paths multi-path coefficients between q transmitting antenna to the i reception antenna;

Every transmit antennas of serving as reasons

The frequency domain channel (q transmitting antenna to the i reception antenna frequency domain channel coefficient during n OFDM symbol) that individual training pilot tone is estimated according to steps A.W _qFor relevant with the q transmitting antenna

The dimension transformation matrix, the capable v column element of its u is:

${\left[W_q\right]}_{u,v}=e^{\frac{-\mathrm{<figure-callout\; id="2"\; label="j"\; filenames="C20041002186400182.png"\; state="\{\{state\}\}">j</figure-callout>}2\mathrm{\&pi;}}{N}(q-1+(u-1)D_{\&Integral;0})(v-1)}$

Wherein q is the transmitting antenna sequence number, 1≤q≤M _T, $u=\mathrm{1,2},\&CenterDot;\&CenterDot;\&CenterDot;,\stackrel{\&OverBar;}{P},$ v＝1，2，…，N。

Step C: obtain channel master multipath component time of delay and extract main multipath component fading coefficients and utilize result of calculation among the step B, select (STC) method according to main multipath component, single-shot list between the extremely any reception antenna i of any transmitting antenna q is received (SISO) channel, from its channel multi-path fading coefficients

(wherein 0≤l≤N-1) middle selection

Bigger

Bar master multipath writes down its time of delay (being unit with the sampling period) and is

(between q transmitting antenna to the i reception antenna), the fading coefficients of respective paths

Main multipath component is selected in this example $\stackrel{\&LeftRightArrow;}{L}=6.$

Stage 2: adaptive channel is followed the trail of during the transfer of data:

When receiving data transmission OFDM symbol (n=1,2 ..., N _T-1), to arbitrary reception antenna i wherein, channel estimation methods following (all reception antennas all need repetition following A, B and three steps of C):

Steps A: adaptive channel track algorithm initial value is set:

All transmitting antennas channel estimation results (should be stored in the receiving terminal memory during practical operation) between reception antenna i is designated as following vector:

Wherein

Be in n the OFDM symbol, the vector that the estimated value of SISO channel master multipath fading coefficient is formed between q transmitting antenna to the i reception antenna.When the receiver processing, during the 1st data transmission OFDM symbol (n=1),

Step C by the stage 1 obtains; In other the data transmission OFDM symbol (n=2,3 ..., N _t-1), the value of (7) formula is by channel self-adapting tracking result during n-1 the OFDM symbol is obtained.At this moment, the channel estimating initial value being set is the final value that a last OFDM symbol is estimated:

Wherein

Represent the estimated result after p iteration of i reception antenna process in n the OFDM symbol,

Be that corresponding n OFDM symbol self adaptation estimate final result, i.e. result after adaptive algorithm iteration P time.

Step B: adaptive channel is followed the tracks of:

Here adopt the LMS algorithm to realize the self-adoptive trace of channel.

To p arbitrarily (p=0,1 ..., P-1) (promptly in any data transmission OFDM symbol) to all reference pilot symbols subcarriers:

LMS algorithm reference signal is Y _i[n, k _p] (received signal at p reference pilot subcarrier of i reception antenna place during n OFDM symbol), the input signal vector of adaptive algorithm is obtained in conjunction with reference pilot sub-carrier positions, main multipath component by known reference pilot signal time of delay, and production method is:

The input signal vector w of LMS algorithm when the p time iteration _p[n] is configured to:

$w_p\left[n\right]={\left[{\mathrm{<figure-callout\; id="1"\; label="w"\; filenames="C20041002186400171.png,C20041002186400182.png"\; state="\{\{state\}\}">w</figure-callout>}}_{1,p}^T\right[n],{\mathrm{<figure-callout\; id="2"\; label="w"\; filenames="C20041002186400182.png"\; state="\{\{state\}\}">w</figure-callout>}}_{2,p}^T[n],\&CenterDot;\&CenterDot;\&CenterDot;,w_{M_T,p}^T[n\left]\right]}^T$

Wherein

The value of representing p reference pilot symbols of q transmitting antenna in n the data transmission OFDM, receiving terminal is known should to be worth;

Step C by the stage 1 selects.In the current data transmission OFDM symbol, iteration P=64 time altogether obtains

As the time domain channel results estimated in this OFDM symbol.

The method of the realization channel tracking of LMS algorithm is:

To all p=0,1 ..., P-1, iterative computation:

$e[n,p]=Y_i[n,k_p]-w_p^T\left[n\right]{\tilde{h}}_i[n,p]$

${\tilde{h}}_i[n,p+1]={\tilde{h}}_i[n,p]+2\mathrm{\&mu;}{w}_p^{*}\left[n\right]e[n,p]$

Wherein: μ is a step factor, selects foundation for to make algorithmic statement and to have less evaluated error, and LMS algorithm step factor is taken as μ=0.02 in this example.

Step C: utilize step B adaptive channel results estimated to estimate all channel sub-carrier frequencies responses

By

Fractionation obtains all transmitting antennas being estimated as to each the main multipath component of channel between reception antenna i

Wherein

Be the estimated value (is unit with the sampling period) of this main multidiameter time, $\stackrel{\&LeftRightArrow;}{l}=\mathrm{1,2},\&CenterDot;\&CenterDot;\&CenterDot;,\stackrel{\&LeftRightArrow;}{L}.$ Then through obtaining during n the OFDM symbol channel k sub-carrier frequency domain channel estimating between q transmitting antenna to the i reception antenna behind the Fourier transform For:

Wherein k represents that (its span is 0≤k＜N-1) for the sequence number of subcarrier.This routine intermediate frequency spectrum loss in efficiency is $\mathrm{\&xi;}=\frac{N+(D_i-1)P}{D_{i}N},$ Promptly 8.6%.Fig. 6 has provided bit error rate (BER) the computer policy performance curve that uses the present invention program's a special case, and simulation result shows that than high s/n ratio the time, the snr loss is about 1dB in the present embodiment.

Claims (1)

1. the adaptive channel estimation method of a MIMO-OFDM system is characterized in that it comprises make a start step and receiving end step:

The described step of making a start is carried out according to step 1 and step 2:

Step 1: introduce training frequency guide symbol and reference pilot symbols to form data frame structure

To any transmitting antenna, transmit data according to a minute frame mode, every frame comprises N _tIndividual OFDM symbol, N _tBe positive integer; First OFDM symbol is a training OFDM symbol in the frame, and n=0, n represent OFDM symbol sequence number; N behind the accent white silk OFDM symbol in the frame _t-1 OFDM symbol is a data transmission OFDM symbol, n=1, and 2 ..., N _t-1; Insert training frequency guide symbol and insert reference pilot symbols according to following method:

Insert training frequency guide symbol: in training OFDM symbol, for q transmitting antenna, q=1,2 ..., M _T, M _TThe expression number of transmit antennas, M _TBe positive integer, first subcarrier number that transmits training frequency guide symbol is q-1, thereafter every D _F0Individual subcarrier inserts a training frequency guide symbol; To q transmitting antenna, at non-training frequency guide symbol subcarrier place, the transmission amplitude is 00 symbol;

Insert reference pilot symbols: in any data transmission OFDM symbol, all transmitting antennas use identical subcarrier to transmit reference pilot symbols; To any transmitting antenna, the subcarrier sequence number that transmits the 1st reference pilot symbols is k ₀, 0≤k ₀＜D _F1, D _F1Be positive integer, thereafter every D _F1Individual subcarrier inserts a reference pilot symbols; To any transmitting antenna, at non-reference pilot symbols subcarrier place, transmit the data symbol of Space Time Coding output: the subcarrier sequence number that transmits reference pilot symbols is k _p, p=0,1 ..., P-1,

N is the OFDM sub-carrier number;

Step 2: the frame structure that forms according to step 1 sends.

Described receiving end step was carried out according to following stage 1 and stage 2:

Stage 1: estimate channel at the training OFDM symbol place of every frame, be used for the initial parameter that adaptive channel is followed the trail of during the transfer of data in this frame, carry out in turn according to A, B and three steps of C to obtain:

Steps A: the estimated value of obtaining the frequency response of all training frequency guide symbol subcarriers of current channel with least-squares estimation;

Step B: to the result of steps A, use frequency domain-＞spatial transform estimates that all transmitting antennas are to channel impulse response between all reception antennas, if in the channel impulse response, being estimated as of l bar multipath component multipath fading coefficient between q transmitting antenna to the i reception antenna in n OFDM symbol 0≤l≤N-1, i=1,2 ..., M _R, M _RBe the reception antenna number;

Step C: obtain channel master multipath component time of delay and extract main multipath component fading coefficients:

Utilize result of calculation among the step B, use direct truncation method of channel impulse response or main multipath component back-and-forth method, to the channel between the extremely any reception antenna i of any transmitting antenna q, from its channel multi-path fading coefficients

The middle selection

Bar master multipath,

Be positive integer, storage delay time l _{I, q, k}Fading coefficients with respective paths

$k=\mathrm{1,2},\&CenterDot;\&CenterDot;\&CenterDot;,\stackrel{\&infin;}{L};$

Stage 2: adaptive channel is followed the trail of during the transfer of data:

To n data transmission OFDM symbol in the frame, n=1,2 ..., N _t-1, wherein arbitrary reception antenna i, use following channel estimation methods:

Steps A: adaptive channel track algorithm initial value is set:

Adaptive channel is set estimates that initial value is the final value that n-1 OFDM symbol estimated:

n＝1，2，…，N _t-1 (1)

Wherein:

Represent in n the OFDM symbol that through after p iteration, all transmitting antennas are to the estimated result of channel between the reception antenna i,

Represent in n the OFDM symbol that through after P iteration, all transmitting antennas are to the final result of channel estimating between the reception antenna i,

Represent in n the OFDM symbol that through after P iteration, q transmitting antenna is to the final result of channel estimating between the reception antenna i, wherein

Step B: adaptive channel is followed the tracks of:

Arbitrarily the reference pilot symbols subcarrier is all corresponding to an iteration of adaptive algorithm, and to p arbitrarily, 0≤p≤P-1, P are the reference pilot symbols number, and the reference signal of adaptive algorithm is Y _i[n, k _p], the received signal at p reference pilot subcarrier of i reception antenna place during n OFDM symbol, k _pFor transmitting the subcarrier sequence number of p reference pilot symbols;

The input signal vector w of adaptive algorithm _p[n] is:

$w_p\left[n\right]=\left[w_{1.p}^T\right[n],w_{2.p}^T[n],...,w_{M_T.p}^T{\left[n\right]]}^T---\left(4\right)$

Wherein

The value of representing p reference pilot symbols of q transmitting antenna in n the data transmission OFDM, receiving terminal is known should to be worth;

Adaptive channel is estimated: adopt the adjustment of LMS algorithm or RLS algorithm or QR-RLS algorithm

In the current data transmission OFDM symbol, iteration is carried out P time altogether, and P is the number of reference pilot symbols in each OFDM symbol, obtains As the time domain channel results estimated in the OFDM symbol, promptly

Step C: utilize step B adaptive channel results estimated to estimate the order of all channel sub-carrier frequencies responses, split according to the combination of (2) and (3) formula

Obtain of the estimation of all transmitting antennas to each the main multipath component of channel between reception antenna i; The estimated value of remembering main multipath component between q transmitting antenna to the i reception antenna is

Use to obtain during n the OFDM symbol channel k sub-carrier frequency domain channel estimating between q transmitting antenna to the i reception antenna behind the Fourier transform,

All reception antennas are repeated above-mentioned steps A, B and C.

Images