CN1491053A - Space-time joint detection device and method based on discrete Fourier transform in wireless transmission - Google Patents

Abstract

The space-time joint detection device and method based on discrete Fourier transform in wireless transmission relates to an equalization method and device applied in the field of wireless mobile communication. The detection method of the device is as follows: a. The estimated channel impulse response sequence Perform space-time combination to obtain a new data sequence h(n); b. Use the estimated channel impulse response sequence and the received data sequence to perform space-time combination to obtain a new data sequence r ^∧ (n); c. For r ^∧ (n) performs non-normalized discrete Fourier transform to obtain the transformation coefficient R ^∧ (k); d. conduct discrete Fourier transform to h(n), and take the real part H(k) of the transformed coefficient; e. use H(k), R ^∧ (k) and the estimated channel noise variance are processed by an equalizer to obtain a new data sequence S ^∧ (k); f. Perform unnormalized inverse discrete Fourier on the sequence S ^∧ (k) Transform to get the reconstructed sent data sequence S ^∧ (n).

Description

Space-time joint detection device and method based on discrete Fourier transform in wireless transmission

1. Technical field:

The present invention relates to an equalization method and device applied in the field of wireless mobile communication, in particular to a method and device for equalization of a wireless spread spectrum communication system under fading channel conditions.

2. Background technology:

In the mobile communication transmission system, due to the multipath propagation of the wireless channel, it is very easy to cause intersymbol interference due to the influence of its channel characteristics. When the interference is seriously affected, the system must be corrected. The equalization technology is mainly to resist multipath fading. interference. In a digital communication system, if the channel frequency response within the system bandwidth is non-flat, then in the time domain, the impulse response of the channel will spread with time delay. This time delay spread is also called channel dispersion, and all time-division systems with channel dispersion can use equalizers to reduce the system performance degradation caused by intersymbol interference.

There are two types of equalization: time domain equalization and frequency domain equalization. Time-domain equalization can be divided into linear equalization and nonlinear equalization. Frequency domain equalization is to connect a filter in series at the receiving end to compensate the amplitude-frequency and phase-frequency characteristics of the channel.

There are three commonly used linear equalization algorithms: the zero-forcing algorithm based on the minimum peak distortion criterion, the MMSE algorithm based on the minimum mean square error criterion, and the least squares algorithm based on the minimum error sum of squares criterion. In addition, these three algorithms have their non-linear variants—decision feedback equalization (DFE) algorithm, but the most commonly used one is the DFE algorithm based on MMSE criterion. For situations where the channel parameters are unknown or the channel is time-varying, the above equalization algorithms have their own adaptive algorithms: adaptive zero-forcing algorithm, least mean square (LMS) algorithm, recursive least squares (RLS) algorithm and adaptive DFE algorithm.

Although the zero-forcing algorithm has a small amount of calculation, its performance is poor, and it has the limitation of peak distortion, so its application is greatly limited. Although the calculation amount of the LMS algorithm is small, the performance is better when the channel frequency response is relatively flat, but the performance becomes worse when the channel response is not flat. The RLS algorithm is characterized by fast convergence speed and high convergence precision, but the computational complexity is relatively high. The characteristic of decision feedback equalization DFE is that if the previous decision is correct, then the feedback filter can completely remove the interference from the previous symbol. However, if a decision is wrong, it may lead to error propagation.

Another class of non-linear equalizer algorithms with detection error probability as the optimization criterion. The most typical of such algorithms are maximum likelihood sequence estimation (MLSE) algorithm and maximum a posteriori probability (MAP) algorithm.

The MAP algorithm also has two disadvantages: one is that it needs to perform forward and backward recursive operations, and since the recursive process contains a large number of exponential summation operations, the computational overhead is very large; second, the decision delay of the MAP algorithm Larger, which leads to a large storage requirement. For the long frame transmission system, the huge storage space is extremely unfavorable for reducing the hardware overhead of the receiver.

The MLSE algorithm is usually implemented by the classic Viterbi algorithm (VA). The state number of VA is related to the register, which is usually suitable for small memory length. In addition, regardless of the size of the channel interference, the calculation amount of each selected path is constant, so when the channel interference is small, the average calculation amount of decoding is too large

Frequency domain equalization is to connect a filter in series at the receiving end to compensate the amplitude-frequency and phase-frequency characteristics of the channel. For example, frequency domain equalization is used in Orthogonal Frequency Division Multiple Access (OFDM) systems to estimate the frequency response of the channel. Due to the use of discrete Fourier transform (DFT) to achieve modulation and demodulation, it is more sensitive to carrier frequency offset, phase noise and nonlinear amplification. Avoiding signal distortion and spectral spreading requires a linear amplifier with a large dynamic range. The basis of OFDM is that the sub-carriers must meet the requirements of orthogonality. If the orthogonality deteriorates, the performance of the entire system will be severely degraded, resulting in OFDM-specific inter-sub-carrier crosstalk. In actual work, due to the time-varying nature of the wireless fading channel, frequency dispersion is often caused, causing Doppler frequency shift effect, thus affecting the carrier frequency orthogonality.

3. Contents of the invention:

1. Technical issues

The object of the present invention is to overcome the aforementioned problems and provide a space-time joint detection device and method based on discrete Fourier transform in wireless transmission that reduces the required calculation complexity.

2. Technical solution

The space-time joint detection device based on discrete Fourier transform in the wireless transmission of the present invention is characterized in that the device includes a local pilot unit, a pilot data branch, a channel estimation unit, a time-division combining unit, a time-space combining unit, and an averaging unit, A discrete Fourier transform unit, an equalizer, and a despreading and deinterleaving unit; wherein the output ends of the local pilot unit and the pilot data branch are respectively connected to the input ends of the channel estimation unit, and the output ends of the channel estimation unit are respectively connected to time-space combining and averaging unit, and correspondingly connected with the time-division merging unit, the output terminal of the time-division merging unit is connected to the input end of the space merging unit, the output terminal of the space merging unit is connected to the input end of the discrete Fourier transform unit, and the output terminal of the space-time merging unit is connected to the discrete Fourier transform unit The input terminal of the average unit, the output terminal of the discrete Fourier transform unit is connected to the input terminal of the equalizer, the output terminal of the equalizer is connected to the input terminal of the discrete Fourier transform unit, and the output terminal of the diffuse Fourier transform unit is connected to the despreading deinterleaving unit The input end and the output end of the despreading and interleaving unit are the output ends of the device, and the input end of the local pilot unit is connected with timing signals.

The space-time joint detection method of the discrete Fourier transform in the block wireless transmission is as follows:

a. Perform space-time combination on the estimated channel impulse response sequence to obtain a new data sequence h(n);

b. Using the estimated channel impulse response sequence and the received data sequence to perform space-time combination to obtain a new data sequence r^(n);

c, carry out unnormalized discrete Fourier transform to r^(n), obtain transformation coefficient R^(k);

D, carry out discrete Fourier transform to h (n), and get the real part H (k) of the transformed coefficient;

e. Use H(k), R^(k) and the estimated channel noise variance to obtain a new data sequence S^(k) through equalizer processing;

f, unnormalized inverse discrete Fourier transform (IDFT) is carried out to sequence S^(k), obtains the transmitted data sequence s^(n) of reconstruction;

In step a of the above-mentioned method:

(1) Each estimated channel impulse response sequence is a 6-path channel parameter;

(2) Combine the estimated channel impulse response sequence in space-time to obtain a 6-point sequence h(n);

(3) h(n) is a sequence of conjugate symmetry about the origin;

(4) Channel estimation is a channel estimation method based on cyclic orthogonal pilot sequences

(5) In the channel estimation, the least squares method is used to obtain the first channel estimation;

(6) Obtaining the second channel estimation by adopting polynomial fitting method to the first channel estimation;

(7) Estimate the noise variance by using the channel estimate obtained for the second time.

(8) The above-mentioned channel estimation is carried out for each pilot frequency segment;

(9) After obtaining the second channel estimation of each pilot frequency segment, the channel estimation of each sub-slot can be obtained by the average of the second channel estimation of the two adjacent pilot frequency segments before and after the data and control information segment;

(10) The noise variance of each sub-slot can be obtained by the average of the channel noise variance estimated on the two pilot segments adjacent to the data and control information segments;

The time slot structure adopted by the sending end is composed of one or more sub-slots, and each time slot is composed of one or more sub-slots (G+P+D&C) and a tail, in which there are cycle protection (G), pilot ( P), user data (D), control information (C), data and control information segment (D&C), trailer (G+P).

The method of space-time merging is:

First correlate the 6-path channel parameters estimated for each sub-slot of each antenna to obtain a new 6-point sequence;

Combine the new 6-point sequences of the corresponding sub-slots of the 4 antennas to obtain a new sequence h(n);

Perform transformation h(n)=2*h(n)(n≠0)

The method of space-time merging is:

a. First, use the estimated 6-point channel impulse response sequence for each sub-slot of each antenna to perform maximum ratio combination on the received data sequence;

b. add the 4 sequences respectively corresponding to 4 antennas after the above-mentioned merging, and obtain a new data sequence r^(n), with a length of _Ld , where _Ld is the length of the transmitted data sequence;

Perform non-normalized discrete Fourier transform on r^(n), and the steps to obtain the transformation coefficient R^(k) are:

a. The DFT of unnormalized r^(n) is L _d point;

b. Non-normalization means that when the L _d point DFT transformation sequence is accumulated and summed, its value is not unit 1.

h(n) is carried out to discrete Fourier transform, and when taking the real part H(k) of the transformed coefficient, the DFT transformation carried out to h(n) is L _d point;

When H(k), R^(k) and the estimated channel noise variance are processed by an equalizer to obtain a new data sequence S^(k):

The equalizer used is a minimum mean square error (MMSE) equalizer;

Equalization is performed at a single point in the DFT domain;

The equalization process is: R^(k)/(H(k)+N ₀ ), N ₀ is the estimated noise variance;

The length of the sequence S^(k) is L _d .

When the unnormalized inverse discrete Fourier transform (IDFT) is performed on the sequence S^(k) to obtain the reconstructed transmitted data sequence s^(n), the unnormalized means that when the adopted IDFT transform sequence is accumulated and summed, its Value is not unit 1.

The equalization method is also characterized by:

The single-point equalization method based on DFT domain is carried out in a sub-slot.

In the above device, the pilot/data splitting unit is used to split the received data; the channel estimation part adopts the channel estimation method based on the circular orthogonal pilot sequence. First, the first channel estimation of least squares is performed on each pilot segment, and after the channel estimation of each pilot segment of the entire time slot is obtained, the second channel estimation is obtained by using the polynomial fitting method for the estimated parameters; The second channel estimate estimates the noise variance; the second channel estimate and noise variance estimate are sent to subsequent units for use. The local pilot unit is used to store the local pilot sequence; the MMSE equalizer unit is performed at a single point in the frequency domain.

3. Beneficial effects

The DFT domain single-point equalization scheme shown in Figure 1. The main cost of implementation is 2 complex DFTs and 1 complex inverse DFT (IDFT). At the same time, by using the short channel impulse response (estimating 6-path channel parameters), the cost of one DFT operation can be greatly reduced, which is beneficial to hardware implementation and improved operation speed. The simulation shows that: when the signal-to-noise ratio is 6-12dB, the bit error rate performance of this system is more than an order of magnitude better than that of the OFDM system by using the single-point equalization joint detection method in the DFT domain. Refer to Figure 5

4. Description of drawings:

Fig. 1 is a schematic diagram of an intermittent pilot time slot structure adopted for a low-speed moving object. Among them are cycle protection G, pilot frequency P, user data D, and control information C.

Fig. 2 is a schematic diagram of an intermittent pilot time slot structure adopted for a medium-speed moving object.

Fig. 3 is a schematic diagram of an intermittent pilot time slot structure adopted for a high-speed moving object.

Fig. 4 is a specific device block diagram of the present invention. There are local pilot unit 1, pilot data branching 201, 202, 203, 204, channel estimation unit 301, 302, 303, 304, time division combining unit 401, 402, 403, 404, space-time combining 503, and averaging unit 502 , a spatial combining unit 501 , discrete Fourier transform units 601 , 602 , and 603 , an equalizer 701 , and a despreading and deinterleaving unit 801 .

Fig. 5 is a simulation experiment result diagram of the bit error performance of the single-point equalization space-time joint detector in the DFT domain.

Fig. 6 is an application diagram of the present invention in a baseband transmission system.

5. Specific implementation methods:

The space-time joint detection device of the discrete Fourier transform in the block wireless transmission of the present invention is characterized in that the device includes a local pilot unit 1, pilot data branches 201, 202, 203, 204, and channel estimation units 301 and 302 , 303, 304, time division merging unit 401, 402, 403, 404, time-space merging 503, averaging unit 502, discrete Fourier transform unit 601, 602, 603, equalizer 701, despreading and deinterleaving unit 801; wherein the local pilot The output terminals of the unit 1 and the pilot data branch 201, 202, 203, 204 are respectively connected to the input terminals of the channel estimation units 301, 302, 303, 304, and the output terminals of the channel estimation units 301, 302, 303, 304 are respectively connected to the space-time Combining 503, averaging unit 502, and correspondingly connecting with time division merging unit 401, 402, 403, 404, the output terminal of time division merging unit 401, 402, 403, 404 is connected to the input end of space merging unit 501, and space merging unit 501 The output terminal of the output terminal is connected to the input terminal of the discrete Fourier transform unit 602, the output terminal of the space-time combination 503 is connected to the input terminal of the discrete Fourier transform unit 601, and the output terminal of the averaging unit 502 and the discrete Fourier transform unit 601,602 are connected to the input of the equalizer 701 terminal, the output terminal of the equalizer 701 is connected to the input terminal of the discrete Fourier transform unit 603, the output terminal of the discrete Fourier transform unit 603 is connected to the input terminal of the despreading deinterleaving unit 801, and the output terminal of the despreading deinterleaving unit 801 is the The output terminal and the input terminal of the local pilot unit 1 are connected with a timing signal.

In order to make the purpose of the present invention, technical solutions and advantages clearer, the implementation of the technical solutions will be further described in detail below in conjunction with the accompanying drawings:

Figure 1, Figure 2 and Figure 3 show the time slot structures used for slow, medium and high-speed moving objects, respectively. Each time slot is 0.825 milliseconds long and divided into 1056 chips. Each time slot consists of one or more sub-slots (D&C+P+G) and a tail. In the figure, G represents cycle protection, P represents pilot frequency, D represents user data, and C represents control information. In the figure, each G is composed of 8 chips, and each P is composed of 24 chips. The difference is that D&C in Fig. 1 has 992 chips, and each D&C in Fig. 2 has 480 chips. Fig. 3 Each D&C in has 224 chips.

(1) Fig. 4 is a device block diagram of the above-mentioned channel estimation method and DFT domain single-point equalization method, in which one antenna is used to transmit and 4 antennas are received in the device figure,

The following is a structural block diagram to describe the equalization method. The steps are as follows:

(1) Channel estimation part:

The pilot sequences P _m (n) output from the pilot/data branches 201, 202, 203, 204 and the output of the local pilot sequence unit 1 are sent to the channel estimation units 301, 302, 303, 304 respectively. The channel estimation part adopts the channel estimation method based on the cyclic orthogonal pilot sequence as mentioned above. Each channel estimation unit estimates the 6-path channel parameter sequence h ₁ (n), h ₂ (n), h ₃ (n), h ₄ (n) and the corresponding channel noise variance N ₀₁ , N ₀₂ , N ₀₃ , N ₀₄ .

The block diagram of the structure adopts the way of parallel processing of four antennas, and the way of time division multiplexing can also be used in the process of realization, that is, the serial way of processing one antenna at a time.

(2) Space-time combination of estimated channel parameters

The obtained 6-path channel parameter sequences h ₁ (n), h ₂ (n), h ₃ (n), h ₄ (n) are sent to space-time combining module 1 for space-time combining to obtain a 6-point sequence $h\left(\mathrm{no}\right)=\underset{m=1}{\overset{4}{\mathrm{\Σ}}}\underset{l}{\mathrm{\Σ}}{h}_m^{*}\left(l\right)h_m(\mathrm{no}+l),$ and do the transformation

h(n)=2*h(n)(n≠0)

(3) Reconstruct the sent data sequence:

The parameter sequence h ₁ (n), h ₂ (n), h ₃ (n), h ₄ (n) of the channel transfer function and the data output of the pilot/data unit 201, 202, 203, 204 are respectively sent to the time merging unit 401, 402, 403, and 404 respectively perform time-combined H ₁ *r ₁ , H ₂ *r ₂ , H ₃ *r ₃ , and H ₄ *r ₄ operations. Where r ₁ , r ₂ , r ₃ , and r ₄ are the data sequences received by each antenna output by the pilot/data unit. H _m (where m=1, 2, 3, 4 correspond to each antenna respectively) is a circular matrix in the following form:

The channel noise variances N ₀₁ , N ₀₂ , N ₀₃ , and N ₀₄ are sent to the averaging unit 305 for averaging.

(4) Carry out space merging on the reconstructed sending data sequence:

The output r ₁ ^(n), r ₂ ^(n), r ₃ ^(n), r ₄ ^(n) of the time merging modules 401, 402, 403, 404 are sent to the space merging unit 501 to obtain the sequence

${\mathrm{<span\; class="notranslate">\; r</span>}}^{<span\; class="notranslate">\; ^</span>}<span\; class="notranslate">\; (</span>\mathrm{<span\; class="notranslate">\; no</span>}<span\; class="notranslate">\; )</span><span\; class="notranslate">\; =</span>{\mathrm{<span\; class="notranslate">\; \&Sigma;</span>}}_{\mathrm{<span\; class="notranslate">\; m</span>}<span\; class="notranslate">\; =</span>\mathrm{<span\; class="notranslate">\; 0</span>}}^{\mathrm{<span\; class="notranslate">\; 4</span>}}{\mathrm{<span\; class="notranslate">\; \&Sigma;</span>}}_{\mathrm{<span\; class="notranslate">\; p</span>}<span\; class="notranslate">\; =</span>\mathrm{<span\; class="notranslate">\; 0</span>}}^{\mathrm{<span\; class="notranslate">\; p</span>}<span\; class="notranslate">\; -</span>\mathrm{<span\; class="notranslate">\; 1</span>}}{\mathrm{<span\; class="notranslate">\; h</span>}}_{\mathrm{<span\; class="notranslate">\; m</span>}}^{<span\; class="notranslate">\; *</span>}<span\; class="notranslate">\; (</span>\mathrm{<span\; class="notranslate">\; p</span>}<span\; class="notranslate">\; )</span>{\mathrm{<span\; class="notranslate">\; r</span>}}_{\mathrm{<span\; class="notranslate">\; m</span>}}<span\; class="notranslate">\; (</span>{<span\; class="notranslate">\; (</span><span\; class="notranslate">\; (</span>\mathrm{<span\; class="notranslate">\; no</span>}<span\; class="notranslate">\; +</span>\mathrm{<span\; class="notranslate">\; p</span>}<span\; class="notranslate">\; )</span><span\; class="notranslate">\; )</span>}_{{\mathrm{<span\; class="notranslate">\; L</span>}}_{\mathrm{<span\; class="notranslate">\; d</span>}}}<span\; class="notranslate">\; )</span><span\; class="notranslate">\; ,</span>$

表示模L_d运算，n＝0，1，...，L_d-1其中L_d为发送数据序列的长度，P为信道多径数(此处取为6)。in,

Indicates a modulo L _d operation, n=0, 1, ..., L _d -1 where L _d is the length of the transmitted data sequence, and P is the number of channel multipaths (6 here).

(5) Carry out DFT transformation to the channel parameters after space-time merging:

The output sequence h(n) of the space-time merging module 1 is sent to the DFT unit 601, which can also be realized by FFT during specific implementation;

(6) Perform DFT transformation on the reconstructed data sequence after space merging:

The output r^(n) of the space merging unit 501 is sent to the DFT unit 602, which can also be realized by FFT during specific implementation;

(7) Single-point equalization in the frequency domain:

Get the real part H(k) of the output coefficient of the DFT unit 601 and the output R^(k) of the DFT unit 602 and the output _N of the averaging unit 305 and send the MMSE equalizer unit 701 to perform single-point equalization R^( in the DFT domain k)/(H(k)+N ₀ );

(8) Acquisition of time domain data sequence:

The output S^(k) of the MMSE equalizer sends the inverse DFT (IDFT) module 603 to obtain the time domain data sequence s^(k); also can adopt IFFT to realize when concrete realization;

(9) Subsequent processing of time domain data sequence

Send the s^(k) sequence to the follow-up module 801 and so on.

Fig. 5 is the simulation experiment result of the bit error performance of the single-point equalization space-time joint detector in the DFT domain. It can be seen from the figure that when the signal-to-noise ratio is 6-12dB, the bit rate performance of this system is more than an order of magnitude better than that of the OFDM system by using the single-point equalization joint detection method in the DFT domain.

Fig. 6 is an application of the present invention in a baseband transmission system.

This equalization method is mainly realized by FPGA in application.

Claims (10)

1. A space-time joint detection device based on discrete Fourier transform in wireless transmission, characterized in that the device includes a local pilot unit (1), pilot data branching (201, 202, 203, 204), channel estimation Unit (301, 302, 303, 304), time division merging unit (401, 402, 403, 404), space-time merging (503), averaging unit (502), space merging unit (501), discrete Fourier transform unit (601 , 602, 603), equalizer (701), despreading and deinterleaving unit (801); wherein the output ends of the local pilot unit (1) and the pilot data branch (201, 202, 203, 204) are respectively connected to the channel The input terminals of the estimation unit (301, 302, 303, 304), the output terminals of the channel estimation unit (301, 302, 303, 304) are respectively connected to the space-time combination (503), the averaging unit (502), and correspond to the time division combination The units (401, 402, 403, 404) are connected, the output terminals of the time division merging units (401, 402, 403, 404) are connected to the input terminals of the space merging unit (501), and the output terminals of the space merging unit (501) are connected to the discrete The input end of the Fourier transform unit (602), the output end of the space-time combination (503) is connected to the input end of the discrete Fourier transform unit (601), the output end of the averaging unit (502), the discrete Fourier transform unit (601,602) is connected The input terminal of the equalizer (701), the output terminal of the equalizer (701) is connected to the input terminal of the discrete Fourier transform unit (603), and the output terminal of the discrete Fourier transform unit (603) is connected to the despreading and deinterleaving unit (801) input end, the output end of the despreading and interleaving unit (801) is the output end of the device, and the input end of the local pilot unit (1) is connected with a timing signal. 2. A detection method suitable for a space-time joint detection device based on discrete Fourier transform in wireless transmission according to claim 1, characterized in that the space-time joint detection method of discrete Fourier transform in the block wireless transmission as follows: a. Perform space-time combination on the estimated channel impulse response sequence to obtain a new data sequence h(n); b. Using the estimated channel impulse response sequence and the received data sequence to perform space-time combination to obtain a new data sequence r^(n); c, carry out unnormalized discrete Fourier transform to r^(n), obtain transformation coefficient R^(k); D, carry out discrete Fourier transform to h (n), and get the real part H (k) of the transformed coefficient; e. Use H(k), R^(k) and the estimated channel noise variance to obtain a new data sequence S^(k) through equalizer processing; f, unnormalized inverse discrete Fourier transform (IDFT) is carried out to sequence S^(k), obtains the transmitted data sequence s^(n) of reconstruction; 3. The space-time joint detection method based on discrete Fourier transform in wireless transmission according to claim 2, characterized in that in step a of the above-mentioned method: (1) Each estimated channel impulse response sequence is a 6-path channel parameter; (2) Combine the estimated channel impulse response sequence in space-time to obtain a 6-point sequence h(n); (3) h(n) is a sequence of conjugate symmetry about the origin; (4) Channel estimation is a channel estimation method based on cyclic orthogonal pilot sequences (5) In the channel estimation, the least squares method is used to obtain the first channel estimation; (6) Obtaining the second channel estimation by adopting polynomial fitting method to the first channel estimation; (7) Estimate the noise variance by using the channel estimate obtained for the second time. (8) The above-mentioned channel estimation is carried out for each pilot frequency segment; (9) After obtaining the second channel estimation of each pilot frequency segment, the channel estimation of each sub-slot can be obtained by the average of the second channel estimation of the two adjacent pilot frequency segments before and after the data and control information segment; (10) The noise variance of each sub-slot can be obtained by the average of the channel noise variance estimated on the two pilot segments adjacent to the data and control information segments; 4. The space-time joint detection method based on discrete Fourier transform in wireless transmission according to claim 2 or 3, wherein the time slot structure adopted by the sending end is composed of one or more sub-slots, and each time slot is composed of One or more sub-slots (G+P+D&C) and tails, which respectively have cycle protection (G), pilot (P), user data (D), control information (C), data and control information segments ( D&C), tail (G+P). 5. The space-time joint detection method based on discrete Fourier transform in wireless transmission according to claim 2, characterized in that the space-time combination method is: (1) First correlate the 6-path channel parameters estimated by each sub-slot of each antenna to obtain a new 6-point sequence; (2) Merge the new 6-point sequences of the corresponding sub-slots of the 4 antennas to obtain a new sequence h(n); (3) Transform h(n)=2*h(n)(n≠0) 6. The space-time joint detection method based on discrete Fourier transform in wireless transmission according to claim 2 or 5, characterized in that the space-time combination method is: (1) First, use the estimated 6-point channel impulse response sequence for each sub-slot of each antenna to perform maximum ratio combination on the received data sequence; (2) adding the 4 sequences respectively corresponding to the 4 antennas after the above-mentioned merging, to obtain a new data sequence r^(n), the length is _Ld , where _Ld is the length of the transmitted data sequence; 7. The space-time joint detection method based on discrete Fourier transform in wireless transmission according to claim 2, characterized in that r^(n) is subjected to non-normalized discrete Fourier transform to obtain transform coefficient R^(k) The steps are: (1) The DFT of unnormalized r^(n) is L _d point; (2) Non-normalization means that when the L _d point DFT transformation sequence is accumulated and summed, its value is not unit 1. 8. The space-time joint detection method based on discrete Fourier transform in wireless transmission according to claim 2, characterized in that h(n) is subjected to discrete Fourier transform, and when the real part H(k) of the transformed coefficient is taken , the DFT transformation carried out on h(n) is L _d point; 9. The space-time joint detection method based on discrete Fourier transform in wireless transmission according to claim 2, characterized in that H(k), R^(k) and the estimated channel noise variance are processed by an equalizer to obtain When a new data sequence S^(k): (1) The equalizer adopted is a minimum mean square error (MMSE) equalizer; (2) Equalization is performed at a single point in the DFT domain; (3) The process of equalization is: R^(k)/(H(k)+N ₀ ), where N ₀ is the estimated noise variance; (4) The length of the sequence S^(k) is L _d . 10. The space-time joint detection method based on discrete Fourier transform in wireless transmission according to claim 2, characterized in that the unnormalized inverse discrete Fourier transform (IDFT) is performed on the sequence S^(k) to obtain the reconstructed When sending the data sequence s^(n), non-normalization means that when the IDFT transformation sequence used is accumulated and summed, its value is not unit 1.

Images