CN104539562A - MIMO-OFDM wideband HF channel estimation method - Google Patents

Abstract

The invention asks for protecting an MIMO-OFDM (Multi-Input Multi-Output Orthogonal Frequency Division Multiplexing) wideband HF channel estimation method. The method comprises the following steps: designing a comb-type pilot frame structure format according to the characteristics of a wideband HF channel to determine the insertion position of a pilot; acquiring channel impulse response on the pilot position by taking a least square method as an estimation criterion through an observation sample of an MIMO-OFDM pilot signal; performing filtering through wavelet transform, tracking the maximal value of the wavelet transform on each scale, finding out a signal portion, and filtering noise to increase the accuracy of channel estimation, wherein for a signal mixed with Gaussian white noise, the amplitude of the wavelet transform decreases along with the increase of a wavelet decomposition scale; and acquiring the channel frequency domain response of a full band by adopting an interpolation algorithm based on transform domain zero padding. The method is low in computation complexity, and high in error estimation performance.

Description

Broadband short wave channel estimation method based on multi-input multi-output orthogonal frequency division multiplexing

Technical Field

The invention relates to a channel estimation method in the field of wireless communication, in particular to a channel estimation method based on a broadband shortwave MIMO-OFDM system.

Background

Short-wave communication refers to radio communication in the frequency band 1.5M-30MHz (wavelength 10M-200M). This band includes the short-wave band (3M-30MHz) and part of the medium-wave band (1.5M-3MHz), so short-wave communication is also called High Frequency (HF) communication. Short-wave propagation distance is long, so that short-wave propagation distance is a main wireless communication means for medium and long-distance communication. Meanwhile, the short-wave communication has the advantages of low cost, simple equipment, flexible communication mode, good survivability, strong confidentiality and the like, so the short-wave communication is widely applied to the fields of military affairs, weather, aviation, emergency rescue and disaster relief and the like.

Modern short wave communication is mainly characterized by 'broadband high speed', belongs to the field of wireless communication, and has a complex and severe channel environment. Because short wave transmission is to transmit information by reflection through the ionosphere, the ionosphere frequently moves rapidly and the height of the reflecting layer changes rapidly, so that the length of a propagation path changes continuously, and phenomena such as Doppler shift and Doppler spread are caused. Short-wave transmission also has a multi-hop phenomenon in the transmission process, and the phenomenon can cause multipath effect, so that the short-wave transmission is more complicated. Therefore, accurate modeling is carried out on the short-wave channel, and a set of channel estimation method suitable for short waves is researched, so that the method is particularly important for researching future short-wave communication.

The OFDM (orthogonal frequency division multiplexing) technology is a parallel multi-carrier modulation technology, which converts a coded serial data stream into parallel data streams, then respectively modulates the parallel data by adopting N subcarriers with equal intervals in frequency, and simultaneously transmits the modulated signals of the N subcarriers after adding, so that the frequency spectrum of each symbol only occupies a small part of the channel bandwidth, therefore, the requirement on the frequency spectrum bandwidth is small, in addition, the OFDM technology can also effectively resist multipath delay, and the frequency selective fading resistance is also strong. The MIMO (multiple input multiple output) system improves the capacity of a system channel through a space diversity technology, greatly improves the gain of the system through a space multiplexing technology, and continuously realizes high-efficiency transmission on the premise of ensuring low error rate. The MIMO-OFDM wireless communication system effectively combines the advantages of both, and therefore, research on channel estimation of the MIMO-OFDM system is important in the field of wireless communication.

Channel estimation of MIMO-OFDM systems can be roughly classified into three types according to the method: blind channel estimation, semi-blind channel estimation, and non-blind channel estimation based on parameters. The blind channel estimation and the semi-blind channel estimation have the disadvantages of low convergence speed, complex algorithm, large calculation amount and relatively less application. The non-blind channel estimation is based on the estimation of the pilot frequency symbol or the training sequence, so the algorithm is simpler and the application in practical application is wider. Non-blind channel estimation techniques can be roughly classified into least square method (LS), minimum mean square error method (MMSE), and linear minimum mean square error method (LMMSE). The least square algorithm is low in complexity and easy to implement, but LS channel estimation belongs to unbiased estimation, so that the LS channel estimation is greatly influenced by noise, and accurate channel statistic information is unlikely to be obtained for the actual environment. The estimation algorithm of Minimum Mean Square Error (MMSE) criterion has a good suppression effect on inter-subcarrier interference and white gaussian noise, so the performance of the MMSE estimation algorithm is better than that of the LS estimation algorithm, but the MMSE algorithm needs matrix inversion operation, and when the number N of subcarriers of the system is increased, the operation amount of the matrix becomes very large. Is difficult to realize in engineering. The linear minimum mean square error channel estimation effect is good, but the operation process also involves calculation of a large number of matrixes and covariance, so the operation complexity is also high, and the approximation can be performed by taking a part with a low order.

Disclosure of Invention

In view of the above disadvantages in the prior art, an object of the present invention is to provide a wideband short-wave channel estimation method for reducing the complexity of channel estimation, and the technical solution of the present invention is as follows: a broadband shortwave channel estimation method based on MIMO OFDM comprises the following steps:

101. at a transmitting end, designing a comb pilot frequency adaptive to a fast channel according to parameter indexes of a broadband short wave channel, determining the insertion position of the comb pilot frequency, inserting the designed comb pilot frequency into the determined position, and transmitting the comb pilot frequency through an antenna;

102. pre-designing, extracting pilot frequency data from each antenna receiving end, sequentially sending the extracted pilot frequency data to a designed channel estimator to generate pilot frequency signals, filtering the pilot frequency signals through the filter bank, outputting the filtered pilot frequency signal data, and then calculating a data sampling value when a subcarrier channel at a pilot frequency position estimates a frequency domain by using a least square method to obtain a frequency domain sample value between a transmitting and receiving antenna pair;

103. transforming the frequency domain sample values obtained between each transmitting and receiving antenna pair into a transform domain by adopting a Discrete Fourier Transform (DFT) method, designing a filter through wavelet transformation, performing low-pass filtering on signals in the transform domain, performing zero filling in the transform domain, and performing Inverse Discrete Fourier Transform (IDFT) on a time domain sequence after zero filling to obtain channel responses at all subcarrier positions;

104. and finally, the channel response estimated in the step 103 is sent to a signal processing unit behind each receiving antenna, and the short wave channel estimation of the whole multi-input multi-output orthogonal frequency division multiplexing system is completed.

Further, in step 101, the wideband short-wave channel model adopts a watts channel model, and a simplified model h (t) of the wideband short-wave watts channel model can be represented as follows:

wherein, a_kIs the relative amplitude of path k; l is the number of paths; f. of_dkIs a Doppler shift;is the initial phase; tau is_kIs the relative path delay on path k; is a shock function; t is time.

Further, the frequency interval D of pilot insertion_f：

In the formula tau_maxFor maximum delay spread of the channel response, T_sIs a mimo ofdm symbol period that does not include a cyclic prefix.

Further, the channel estimator is designed according to the LS least squares method, and assuming that the transmitted signal is X, the channel response is H, the received signal is Y, and the noise interference is σ, the channel model is:

Y＝XH+σ

then obtain the LS channel estimation frequency domain responseComprises the following steps:

<math> <mrow> <msub> <mover> <mi>H</mi> <mo>^</mo> </mover> <mi>LS</mi> </msub> <mo>=</mo> <msup> <mi>X</mi> <mrow> <mo>-</mo> <mn>1</mn> </mrow> </msup> <mi>Y</mi> <mo>=</mo> <msup> <mi>X</mi> <mrow> <mo>-</mo> <mn>1</mn> </mrow> </msup> <mrow> <mo>(</mo> <mi>XH</mi> <mo>+</mo> <mi>&sigma;</mi> <mo>)</mo> </mrow> <mo>=</mo> <mi>H</mi> <mo>+</mo> <msup> <mi>X</mi> <mrow> <mo>-</mo> <mn>1</mn> </mrow> </msup> <mi>&sigma;</mi> </mrow> </math>

whereinFor the estimated value after the estimated channel response of the comb pilot,obtained for estimation

The frequency domain channel response.

Further, in step 103, the time domain channel estimation value of the receiving end is:

${\hat{h}}_{\mathrm{ls},k}=h_k+\frac{n_k}{x_k}$

wherein,is the k-thChannel estimation value h on carrier frequency_kIn order to be the true response of the channel,for the interference introduced by noise, the wavelet denoising process is as follows:

wherein c is_k，nIs the coefficient of the low frequency part after decomposition, d_k，nIs the coefficient of the high-frequency part after decomposition, the value range of l is the number of elements of the sequence, n is the nth element in the sequence, all are scale functions, the decomposed high-frequency part is a noise part, and the decomposed low-frequency part is a signal part.

The invention has the following advantages and beneficial effects:

the invention adopts a broadband short wave Watterson simplified model, greatly reduces the complexity of channel estimation, adopts a comb-shaped pilot frequency frame structure to track a channel, is more suitable for a short wave channel with rapid change, adopts a zero-filling interpolation algorithm in a transform domain, and uses the idea of wavelet transform to carry out filtering so as to obtain more accurate channel estimation.

Drawings

FIG. 1 is a block diagram of a wideband short-wave MIMO-OFDM system in accordance with an embodiment of the present invention;

FIG. 2 is a flow chart of interpolation and filtering according to the present invention;

fig. 3 is a flow chart of channel estimation.

Detailed Description

The invention will now be further elucidated with reference to the following non-limiting embodiment in which the drawing is combined. It should be understood that these descriptions are only illustrative and are not intended to limit the scope of the present invention. Moreover, in the following description, descriptions of well-known structures and techniques are omitted so as to not unnecessarily obscure the concepts of the present invention.

A preferred embodiment of the invention is: referring to fig. 1, a MIMO-OFDM wideband short wave channel estimation method is characterized in that: designing a comb-shaped pilot frequency frame structure format according to the characteristics of a broadband short wave channel, namely determining the insertion position of a pilot frequency; obtaining channel impulse response at a pilot frequency position by using an observation sample of the MIMO-OFDM pilot frequency signal and taking a least square method as an estimation criterion; the wavelet transform is used for filtering, the amplitude of the wavelet transform is reduced along with the increase of the decomposition scale of the wavelet of the signal with the aliasing of the Gaussian white noise, the maximum value of the wavelet transform under each scale is tracked, a signal part is found out, noise is filtered, and therefore the accuracy of channel estimation is improved. The method has low calculation complexity and good estimation error performance; and obtaining the channel frequency domain response of the full frequency band by adopting an interpolation algorithm based on transform domain zero padding. The concrete implementation steps are as follows:

(1) accurately inserting the designed comb-shaped pilot frequency into a corresponding position at a transmitting end according to the actual condition of a broadband short wave channel;

(2) extracting pilot frequency data from each receiving antenna end, sequentially sending all the pilot frequency data to a designed channel estimator, filtering the signals through a designed filter bank, outputting pilot frequency signal data, and calculating a subcarrier channel estimation frequency domain data sampling value at a pilot frequency position by using a least square method;

(3) the frequency domain sample value obtained between each transmitting-receiving antenna pair is subjected to DFT conversion to be in a 'conversion domain' defined by the algorithm, a filter is designed through the wavelet conversion idea, signals are filtered in the conversion domain, then zero filling is carried out in the conversion domain, and then IDFT conversion is carried out on the time domain sequence after zero filling to obtain channel responses at all subcarrier positions.

(4) And finally, the estimated channel response is sent to a signal processing unit behind each receiving antenna, and the functions of equalization, space-time decoding or closed-loop transmission of the whole MIMO-OFDM system are completed. In the step (1), the broadband short wave Watterson simplified model is as follows:

wherein, a_kIs the relative amplitude of path k; l is the number of paths; f. of_dkIs a Doppler shift;is the initial phase; tau is_kIs the relative path delay on path k; is a shock function; t is time.

In the step (1), the pilot insertion frequency interval satisfies:

<math> <mrow> <msub> <mi>D</mi> <mi>f</mi> </msub> <mo>&le;</mo> <mfrac> <msub> <mi>T</mi> <mi>s</mi> </msub> <msub> <mi>&tau;</mi> <mi>max</mi> </msub> </mfrac> </mrow> </math>

τ_maxfor maximum delay spread of the channel response, T_sIs a MIMO-OFDM symbol period that does not include a cyclic prefix.

The method comprises the steps of carrying out common filtering on received signals, sending the signals to a DFT unit, sending the signals to the DFT unit, enabling the number of DFT points to be equal to the number of MIMO-OFDM subcarriers, then respectively carrying out least square estimation on data after DFT, and carrying out interpolation processing on the obtained estimated values.

The design of the wavelet transform filter designed in the invention can adopt the idea of multi-time decomposition. The wavelet denoising process is as follows:

wherein c is_k，nIs the coefficient of the low frequency part after decomposition, d_k，nIs the coefficient of the high frequency part after decomposition. The value range of l is the number of elements in the sequence, and n is the nth element in the sequence.Are all scale functions. The decomposed high frequency part is a noise part, and the low frequency part is a signal part, so the high frequency part is filtered after the decomposition of the above formula.

After the small wave filtering, carrying out zero filling of a transform domain, wherein N is the number of all subcarriers, and N is_PEstimating frequency domain response for LS channel for number of subcarriersPerforming DFT conversion to obtainIn transform domain toAnd (3) carrying out zero filling, and carrying out N-point IDFT again, wherein the transform domain formula is as follows:

<math> <mrow> <msub> <mover> <mi>G</mi> <mo>^</mo> </mover> <mi>P</mi> </msub> <mrow> <mo>(</mo> <mi>p</mi> <mo>)</mo> </mrow> <mo>=</mo> <munderover> <mi>&Sigma;</mi> <mrow> <mi>k</mi> <mo>=</mo> <mn>0</mn> </mrow> <msub> <mi>N</mi> <mrow> <mi>P</mi> <mo>-</mo> <mn>1</mn> </mrow> </msub> </munderover> <msub> <mover> <mi>H</mi> <mo>^</mo> </mover> <mi>P</mi> </msub> <mrow> <mo>(</mo> <mi>k</mi> <mo>)</mo> </mrow> <mi>exp</mi> <mrow> <mo>(</mo> <mo>-</mo> <mi>j</mi> <mfrac> <mrow> <mn>2</mn> <mi>&pi;</mi> </mrow> <msub> <mi>N</mi> <mi>P</mi> </msub> </mfrac> <mi>kp</mi> <mo>)</mo> </mrow> <mo>,</mo> <mn>0</mn> <mo>&le;</mo> <mi>p</mi> <mo>&le;</mo> <msub> <mi>N</mi> <mi>P</mi> </msub> <mo>-</mo> <mn>1</mn> </mrow> </math>

<math> <mrow> <msub> <mover> <mi>G</mi> <mo>^</mo> </mover> <mi>N</mi> </msub> <mrow> <mo>(</mo> <mi>q</mi> <mo>)</mo> </mrow> <mo>=</mo> <mfenced open='{' close=''> <mtable> <mtr> <mtd> <msub> <mover> <mi>G</mi> <mo>^</mo> </mover> <mi>P</mi> </msub> <mrow> <mo>(</mo> <mi>q</mi> <mo>)</mo> </mrow> </mtd> <mtd> <mn>0</mn> <mo>&le;</mo> <mi>q</mi> <mo>&le;</mo> <msub> <mi>N</mi> <mi>P</mi> </msub> <mo>-</mo> <mn>1</mn> </mtd> </mtr> <mtr> <mtd> <mn>0</mn> </mtd> <mtd> <msub> <mi>N</mi> <mi>P</mi> </msub> <mo>-</mo> <mn>1</mn> <mo>&le;</mo> <mi>q</mi> <mo>&le;</mo> <mi>N</mi> <mo>-</mo> <mn>1</mn> </mtd> </mtr> </mtable> </mfenced> </mrow> </math>

<math> <mrow> <msub> <mover> <mi>H</mi> <mo>^</mo> </mover> <mi>N</mi> </msub> <mrow> <mo>(</mo> <mi>k</mi> <mo>)</mo> </mrow> <mo>=</mo> <munderover> <mi>&Sigma;</mi> <mrow> <mi>q</mi> <mo>=</mo> <mn>0</mn> </mrow> <mrow> <mi>N</mi> <mo>-</mo> <mn>1</mn> </mrow> </munderover> <msub> <mover> <mi>G</mi> <mo>^</mo> </mover> <mi>N</mi> </msub> <mrow> <mo>(</mo> <mi>q</mi> <mo>)</mo> </mrow> <mi>exp</mi> <mrow> <mo>(</mo> <mo>-</mo> <mi>j</mi> <mfrac> <mrow> <mn>2</mn> <mi>&pi;</mi> </mrow> <mi>N</mi> </mfrac> <mi>qk</mi> <mo>)</mo> </mrow> <mo>,</mo> <mn>0</mn> <mo>&le;</mo> <mi>k</mi> <mo>&le;</mo> <mi>N</mi> <mo>-</mo> <mn>1</mn> </mrow> </math>

where k denotes the frequency domain index and p, q denote the transform domain symbols.I.e. representing all sub-carrier responses.

As shown in fig. 1, a to-be-transmitted bit data stream is transformed into a modulated signal through constellation point mapping, then MIMO encoding is performed to form multiple channels of signals, each channel of signals is OFDM modulated, the modulated signal is transmitted after digital-to-analog conversion and intermediate frequency modulation, a short wave channel is used to receive a signal through a receiver, OFDM demodulation is performed on the received signal, and then MIMO decoding and constellation point mapping are performed on each channel of signals to obtain a transmitted signal.

As shown in FIG. 2, the pilot subchannel response is taken as N_PAnd (pilot frequency subcarrier number) point DFT conversion, filtering in a conversion domain, filtering noise, then zero filling, and then performing N (total subcarrier number) point IDFT conversion to obtain all subcarrier responses.

FIG. 3 is a flow chart of the present invention;

the above examples are to be construed as merely illustrative and not limitative of the remainder of the disclosure. After reading the description of the invention, the skilled person can make various changes or modifications to the invention, and these equivalent changes and modifications also fall within the scope of the invention defined by the method claims.

Claims (5)

1. A broadband shortwave channel estimation method based on MIMO OFDM is characterized in that: the method comprises the following steps:

101. at a transmitting end, designing a comb pilot frequency adaptive to a fast channel according to parameter indexes of a broadband short wave channel, determining the insertion position of the comb pilot frequency, inserting the designed comb pilot frequency into the determined position, and transmitting the comb pilot frequency through an antenna;

102. pre-designing, extracting pilot frequency data from each antenna receiving end, sequentially sending the extracted pilot frequency data to a designed channel estimator to generate pilot frequency signals, filtering the pilot frequency signals through the filter bank, outputting the filtered pilot frequency signal data, and then calculating a data sampling value when a subcarrier channel at a pilot frequency position estimates a frequency domain by using a least square method to obtain a frequency domain sample value between a transmitting and receiving antenna pair;

103. transforming the frequency domain sample values obtained between each transmitting and receiving antenna pair into a transform domain by adopting a Discrete Fourier Transform (DFT) method, designing a filter through wavelet transformation, performing low-pass filtering on signals in the transform domain, performing zero filling in the transform domain, and performing Inverse Discrete Fourier Transform (IDFT) on a time domain sequence after zero filling to obtain channel responses at all subcarrier positions;

104. and finally, the channel response estimated in the step 103 is sent to a signal processing unit behind each receiving antenna, and the short wave channel estimation of the whole multi-input multi-output orthogonal frequency division multiplexing system is completed.

2. The method of claim 1, wherein the method comprises: in step 101, a wideband short-wave channel model adopts a Watterson channel model, and a simplified model h (t) of the wideband short-wave Watterson model can be expressed as follows:

wherein, a_kIs the relative amplitude of path k; l is the number of paths; f. of_dkIs a Doppler shift;is a first

Starting a phase; tau is_kIs the relative path delay on path k; is a shock function; t is time.

3. The MIMO-OFDM based wideband short wave channel of claim 1An estimation method, characterized by: frequency spacing D of pilot insertion_f：

In the formula tau_maxFor maximum delay spread of the channel response, T_sIs a loop not including

A multiple-input multiple-output orthogonal frequency division multiplexing symbol period of the prefix.

4. The method of claim 1, wherein the method comprises: the channel estimator is designed according to the LS least square method, and assuming that a transmitted signal is X, a channel response is H, a received signal is Y, and noise interference is σ, a channel model is:

Y＝XH+σ

then obtain the LS channel estimation frequency domain responseComprises the following steps:

<math> <mrow> <msub> <mover> <mi>H</mi> <mo>^</mo> </mover> <mi>LS</mi> </msub> <mo>=</mo> <msup> <mi>X</mi> <mrow> <mo>-</mo> <mn>1</mn> </mrow> </msup> <mi>Y</mi> <mo>=</mo> <msup> <mi>X</mi> <mrow> <mo>-</mo> <mn>1</mn> </mrow> </msup> <mrow> <mo>(</mo> <mi>XH</mi> <mo>+</mo> <mi>&sigma;</mi> <mo>)</mo> </mrow> <mo>=</mo> <mi>H</mi> <mo>+</mo> <msup> <mi>X</mi> <mrow> <mo>-</mo> <mn>1</mn> </mrow> </msup> <mi>&sigma;</mi> </mrow> </math>

whereinFor the estimated value after the estimated channel response of the comb pilot,is the estimated frequency domain channel response.

5. The method of claim 1, wherein the method comprises: in step 103, the time domain channel estimation value of the receiving end is:

${\hat{h}}_{\mathrm{ls},k}=h_k+\frac{n_k}{x_k}$

wherein,is a channel estimate on the kth carrier frequency, h_kIn order to be the true response of the channel,for the interference introduced by noise, the wavelet denoising process is as follows:

wherein c is_k，nIs the coefficient of the low frequency part after decomposition, d_k，nIs the coefficient of the high-frequency part after decomposition, the value range of l is the number of elements of the sequence, n is the nth element in the sequence, all are scale functions, the decomposed high-frequency part is a noise part, and the decomposed low-frequency part is a signal part.