Ofdm transmission method and system for maritime channels

Abstract

The application discloses an OFDM transmission method and system of a marine channel. The method can include: a transmitting end transmits a signal, converts the signal into a time domain through an IFFT, eliminates interference, converts the signal back into a frequency domain through an FFT, and obtains a frequency domain signal; divides the frequency domain signal into useful data and pilot data, and performs channel estimation on the pilot data; performs equalization, parallel-serial conversion and demodulation on the channel estimated signal and the useful data, and a receiving end obtains a final receiving signal. The application can effectively filter noise interference through an improved channel estimation method.

Description

TECHNICAL FIELD

[0001] The present application relates to the field of wireless communication, more particularly, to an OFDM transmission method and system for maritime channel. BACKGROUND

[0002] Maritime communication is extremely susceptible to the influence of complex marine environment, and it is a difficult problem to ensure high-quality ship communication under the condition of large noise on the sea. Compared with land wireless communication, there are problems such as signal attenuation and signal distortion in maritime wireless communication, and the problem of extremely large noise in maritime communication is more prominent. At present, maritime trade is becoming more and more frequent, accompanied by the continuous increase of various intelligent devices on the ship, and the communication data is also growing rapidly, so the demand for maritime communication is more urgent.

[0003] Orthogonal frequency division multiplexing (OFDM) is a kind of parallel multi-carrier modulation technology, which has been widely used in maritime communication field due to its high spectrum utilization rate and strong anti-interference ability. For OFDM system, LS (traditional least square) algorithm is often used for channel estimation in practice, which is convenient to calculate, but is greatly affected by channel noise, resulting in poor performance. The channel estimation algorithm based on discrete fourier transform (DFT) is an improved algorithm of LS algorithm, which improves the performance of LS algorithm. However, the DFT algorithm only removes the noise outside the cyclic prefix in channel estimation, and the effect is generally improved, which greatly affects the communication quality.

[0004] Therefore, it is necessary to develop an OFDM transmission method and system for maritime channel.

[0005] The information disclosed in the background section of this application is only intended to deepen the understanding of the general background of the application, and should not be regarded as acknowledging or implying in any form that the information constitutes prior art known to those skilled in the art. SUMMARY

[0006] The present application provides an OFDM transmission method and system for maritime channel, which can improve the channel estimation method and effectively filter out noise interference.

[0007] In a first aspect, the present application provides an OFDM transmission method for maritime channel, comprising:

[0008] The transmitting end transmits a signal, which is converted into time domain by IFFT and then converted back into frequency domain by FFT after eliminating interference, to obtain a frequency domain signal;

[0009] The frequency domain signal is divided into useful data and pilot data, and channel estimation is performed on the pilot data;

[0010] The received end obtains final received signals by equalizing and parallel-serial converting and demodulating the channel-estimated signals and the useful data;

[0011] The channel estimation comprises:

[0012] The frequency domain signals are preliminarily estimated by a least square algorithm, and then wavelet packet denoising processing is performed to obtain frequency domain denoising results;

[0013] N-point IDFT is performed on the frequency domain denoising results, all channel responses outside the CP are set to zero, channel response values within the CP length range are reserved, CP internal denoising is completed by improved threshold processing, and time domain denoising processing response values are obtained;

[0014] After N-point DFT conversion is performed on the time domain denoising processing response values, frequency domain channel response values obtained by the conversion are subjected to DFT, the time domain is converted into a transform domain, and filtering data is obtained by filtering the transform domain noise through a low-pass filter;

[0015] IDFT is performed on the filtering data to obtain transform domain denoised frequency domain channel response values, and then a channel-estimated signal is obtained by an interpolation algorithm.

[0016] Preferably, the preliminary estimation of the frequency domain signals by the least square algorithm is:

[0017] The ratio of the received pilot Y to the inserted pilot X is used to obtain the response of the pilot signal in the channel at the transmitting end:

[0018]

[0019] wherein, is the response obtained at the pilot in the frequency domain, H(k) is the frequency response of the channel, and W(k) is the noise in the communication process.

[0020] Preferably, the wavelet packet denoising processing comprises:

[0021] A wavelet is selected and the level to be decomposed is determined, and then wavelet packet decomposition is performed on the signal;

[0022] An optimal wavelet packet basis is defined, an optimal tree is calculated by an entropy standard, and an optimal wavelet basis is obtained;

[0023] A threshold value is determined to perform threshold quantization operation on the wavelet packet decomposition factor;

[0024] A reconstruction operation is performed on the wavelet packet.

[0025] Preferably, N-point IDFT is performed by the following formula:

[0026]

[0027] in, This is the time-domain response after IDFT processing. for After wavelet packet denoising, N is the number of subcarriers, k is the frequency domain interval value, n is the time domain interval value, and j is an imaginary number.

[0028] Preferably, the threshold determination for the improved threshold processing includes:

[0029] Take the median value of the amplitude modulus of all channel responses within the cyclic prefix L as the threshold part:

[0030]

[0031] The threshold portion is the average amplitude of all channel responses excluding the cyclic prefix.

[0032]

[0033] The final threshold is determined by the sum of the two:

[0034] D = α·(D1 + D2)

[0035] Where D is the final threshold, D1 is the median value of the amplitude modulus of all channel responses within the cyclic prefix, D2 is the average value of the amplitude modulus of all channel responses outside the cyclic prefix, α is the multiplication coefficient, median represents the median value, and mean represents the average value.

[0036] Preferably, denoising within the CP is achieved by improving threshold processing:

[0037]

[0038] in, This is the response value after time-domain noise reduction processing.

[0039] Preferably, a low-pass filter is used to filter transform domain noise:

[0040]

[0041] The G(n) index indicates that noise exists throughout the entire transform domain. A low-pass filter is set so that the G(n) corresponding to the high-frequency portion of the transform domain is 0:

[0042]

[0043] in, G(n) is the result after filtering noise, and G(n) is the transform domain response. c It is 1 / 2 the length of the cyclic prefix.

[0044] Preferably, the IDFT is performed on the filtered data using the following formula:

[0045]

[0046] in, This is the frequency domain channel response value after denoising in the transform domain.

[0047] Preferably, the transmitted signal is converted to the time domain by IFFT and interference is eliminated, and then converted back to the frequency domain by FFT to obtain the frequency domain signal, including:

[0048] The signal from the transmitting end is QPSK modulated, and N parallel data channels are obtained by serial-to-parallel conversion in complex form.

[0049] A complex signal X(k) is obtained by inserting a pilot signal; the inserted pilot signal is a block pilot signal.

[0050] The signal is subjected to IFFT to obtain its time-domain form x(n);

[0051] A cyclic prefix is ​​used as the guard interval, and then the parallel-to-serial conversion is performed to send the signal into the channel;

[0052] The received signal is converted from serial to parallel, the cyclic prefix is ​​removed, and then FFT is performed to convert it back to the frequency domain to obtain the frequency domain signal.

[0053] As one specific implementation of this disclosure,

[0054] Secondly, embodiments of this disclosure also provide an OFDM transmission system for maritime channels, comprising:

[0055] The transmitter is used to send signals to the interference removal module;

[0056] The interference removal module is used to convert the signal into the time domain via IFFT and eliminate interference, and then convert it back to the frequency domain via FFT to obtain the frequency domain signal.

[0057] The channel estimation module is used to divide the frequency domain signal into useful data and pilot data, and to perform channel estimation on the pilot data.

[0058] The post-processing module is used to perform equalization, parallel-to-serial conversion, and demodulation on the channel-estimated signal and the useful data, and to send the final received signal to the receiving end.

[0059] The channel estimation module performs channel estimation through the following steps:

[0060] The frequency domain signal is initially estimated using the least squares algorithm, and then the frequency domain denoising result is obtained by wavelet packet denoising.

[0061] An N-point IDFT is performed on the frequency domain denoising result. All external channel responses of the CP are set to zero, while the channel response values ​​within the CP length range are retained. The internal denoising of the CP is completed by improving the threshold processing, and the response value of the time domain denoising is obtained.

[0062] The response value of the time-domain noise reduction process is subjected to an N-point DFT transformation, and then the frequency-domain channel response value obtained by the transformation is subjected to a DFT to convert the domain to the transform domain. The noise in the transform domain is filtered by a low-pass filter to obtain the filtered data.

[0063] The filtered data is subjected to IDFT to obtain the frequency domain channel response value after transform domain denoising, and then the channel-estimated signal is obtained through an interpolation algorithm.

[0064] The methods and systems of the present invention have other features and advantages that will be apparent from or will be set forth in detail in the accompanying drawings and following detailed description, which together serve to explain the particular principles of the invention. Attached Figure Description

[0065] The above and other objects, features and advantages of the present invention will become more apparent from the more detailed description of exemplary embodiments of the invention in conjunction with the accompanying drawings, wherein the same reference numerals generally represent the same parts.

[0066] Figure 1 A flowchart illustrating the steps of an OFDM transmission method for maritime channels is shown.

[0067] Figure 2 A schematic diagram of an OFDM transmission method for a maritime channel according to an embodiment of the present invention is shown. Detailed Implementation

[0068] Preferred embodiments of the invention will now be described in more detail. While preferred embodiments of the invention are described below, it should be understood that the invention can be implemented in various forms and should not be limited to the embodiments set forth herein.

[0069] To facilitate understanding of the solutions and effects of the embodiments of the present invention, two specific application examples are given below. Those skilled in the art should understand that these examples are merely for the purpose of understanding the present invention, and any specific details therein are not intended to limit the present invention in any way.

[0070] Example 1

[0071] Figure 1 A flowchart illustrating the steps of an OFDM transmission method for maritime channels is shown.

[0072] like Figure 1 As shown, the OFDM transmission method for this maritime channel includes: Step 101, the transmitter transmits a signal, which is converted to the time domain by IFFT and interference is eliminated, and then converted back to the frequency domain by FFT to obtain a frequency domain signal; Step 102, the frequency domain signal is divided into useful data and pilot data, and channel estimation is performed on the pilot data; Step 103, the channel-estimated signal and useful data are equalized, converted from parallel to serial, and demodulated, and the receiver obtains the final received signal; Channel estimation includes: preliminary estimation of the frequency domain signal using the least squares algorithm, followed by wavelet packet denoising processing to obtain the frequency domain denoising result; An N-point IDFT is performed on the frequency domain denoising result to set all external channel responses of the CP to zero, retaining the channel response values ​​within the CP length range. Internal denoising of the CP is completed through improved threshold processing to obtain the response value after time-domain denoising. An N-point DFT is then performed on the time-domain denoising response value, followed by another DFT on the resulting frequency domain channel response value to convert the domain to the transform domain. A low-pass filter is used to filter noise in the transform domain to obtain filtered data. An IDFT is then performed on the filtered data to obtain the frequency domain channel response value after transform-domain denoising. Finally, an interpolation algorithm is used to obtain the channel-estimated signal.

[0073] In one example, the initial estimate of the frequency domain signal using the least squares algorithm is as follows:

[0074] At the transmitting end, the response of the pilot signal in the channel is obtained by using the ratio of the received pilot Y to the inserted pilot X.

[0075]

[0076] in, H(k) represents the response obtained at the pilot in the frequency domain, H(k) represents the frequency response of the channel, and W(k) represents the noise during the communication process.

[0077] In one example, wavelet packet denoising includes:

[0078] Select a wavelet and determine the level to be decomposed, then perform wavelet packet decomposition on the signal;

[0079] Define the optimal wavelet packet basis, calculate the optimal tree using the entropy standard, and obtain the optimal wavelet basis;

[0080] Determine the threshold and perform threshold quantization on the wavelet packet decomposition factor;

[0081] Reconstruct the wavelet packet.

[0082] In one example, an N-point IDFT is performed using the following formula:

[0083]

[0084] in, This is the time-domain response after IDFT processing. This is the time-domain response after IDFT processing. for After wavelet packet denoising, N is the number of subcarriers, k is the frequency domain interval value, n is the time domain interval value, and j is an imaginary number.

[0085] In one example, the threshold determination for improved threshold processing includes:

[0086] Take the median value of the amplitude modulus of all channel responses within the cyclic prefix L as the threshold part:

[0087]

[0088] The threshold portion is the average amplitude of all channel responses excluding the cyclic prefix.

[0089]

[0090] The final threshold is determined by the sum of the two:

[0091] D = α·(D1 + D2)

[0092] Where D is the final threshold, D1 is the median value of the amplitude modulus of all channel responses within the cyclic prefix, D2 is the average value of the amplitude modulus of all channel responses outside the cyclic prefix, α is the multiplication coefficient, median represents the median value, and mean represents the average value.

[0093] In one example, denoising within the CP is achieved by improving thresholding:

[0094]

[0095] in, This is the response value after time-domain noise reduction processing.

[0096] In one example, a low-pass filter is used to filter transform domain noise:

[0097]

[0098] The G(n) index indicates that noise exists throughout the entire transform domain. A low-pass filter is set so that the G(n) corresponding to the high-frequency portion of the transform domain is 0:

[0099]

[0100] in, G(n) is the result after filtering noise, and G(n) is the transform domain response.c It is 1 / 2 the length of the cyclic prefix.

[0101] In one example, IDFT is performed on filtered data using the following formula:

[0102]

[0103] in, This is the frequency domain channel response value after denoising in the transform domain.

[0104] In one example, the transmitter transmits a signal, which is then converted to the time domain using IFFT and interference is eliminated. This time domain signal is then converted back to the frequency domain using FFT, yielding the following frequency domain signal:

[0105] QPSK modulation is applied to the signal at the transmitting end, and serial-to-parallel conversion is performed in complex form to obtain N parallel data streams.

[0106] A complex signal X(k) is obtained by inserting a pilot signal; the inserted pilot signal is a block pilot signal.

[0107] The signal is subjected to IFFT to obtain its time-domain form x(n);

[0108] A cyclic prefix is ​​used as the guard interval, and then the parallel-to-serial conversion is performed to send the signal into the channel;

[0109] The received signal is converted from serial to parallel, the cyclic prefix is ​​removed, and then FFT is performed to convert it back to the frequency domain to obtain the frequency domain signal.

[0110] Specifically, AWGN refers to noise during communication. Compared to terrestrial wireless communication, the noise in maritime communication is extremely high. The propagation of wireless signals on the ocean surface is affected by the complex and diverse marine environment, which restricts the propagation of electromagnetic waves on the ocean surface. Large waves at sea further amplify the noise during communication, making the stability of electromagnetic waves highly unstable. With maritime trade becoming increasingly important today, noise control is even more urgent.

[0111] First, the channel response is estimated based on pilot information using the least squares algorithm. Then, wavelet packets are used for noise reduction in the frequency domain. Next, the channel is transformed to the time domain using the inverse discrete Fourier transform and threshold processing is performed in the time domain. Finally, the channel is transformed to the transform domain for further noise reduction.

[0112] For maritime wireless communication transmission systems, selecting a system resistant to multipath fading interference is essential. Compared to other maritime communication methods, OFDM transmission systems can make maritime communication faster and more efficient. A maritime channel model can typically be viewed as a channel with three paths: a direct path, a reflected path, and a scattering path generated by the evaporating waveguide. The reception relationship for a maritime multipath fading channel can be expressed as:

[0113]

[0114] In the formula: p is the number of paths, h p τ is the path fading coefficient. p This refers to path delay.

[0115] The communication process of this multi-domain denoising-based channel estimation algorithm can be summarized into three parts: signal transmission, channel transmission, and signal reception. Among these, the multi-domain denoising-based channel estimation occurs in the signal reception section, and its function is mainly accomplished by wavelet packet denoising, threshold denoising, and transform domain denoising. The specific implementation process is described below.

[0116] Figure 2 A schematic diagram of an OFDM transmission method for a maritime channel according to an embodiment of the present invention is shown.

[0117] like Figure 2 As shown, this method is designed to transmit 300 data symbols per transmission. To enhance anti-interference capability, the input data at the OFDM system transmitter uses QPSK modulation and undergoes serial-to-parallel conversion in complex form to obtain N parallel data streams. The high-speed information data stream after serial-to-parallel conversion is distributed to multiple sub-channels with lower rates for transmission. The symbol period on each sub-channel is also relatively increased, which can reduce the inter-symbol interference caused by multipath delay spread in the wireless channel. A complex signal X(k) is obtained by inserting pilots. In broadband wireless communication systems, the most serious type of interference affecting high-speed information transmission is frequency-selective interference. The inserted pilots are block pilots with a pilot spacing of 5. This pilot method, due to its continuous frequency domain distribution, is suitable for use in frequency-selective fading channels, which are most likely to occur in maritime communications.

[0118] The signal is modulated using an OFDM system via IFFT on multiple orthogonal subcarriers to obtain its time-domain form as x(n). The number of subcarriers is set to 64; orthogonal subcarriers effectively avoid inter-carrier interference, and their spectra can overlap, thus fully utilizing the bandwidth. A cyclic prefix is ​​introduced as a guard interval, with a length of 16. When the cyclic prefix length exceeds the maximum multipath delay spread, inter-symbol interference caused by multipath interference can be eliminated to the greatest extent. The signal is then converted from parallel to serial and sent into the channel. The channel model used is a maritime multipath fading channel with a multipath number of 3. The received signal is converted from serial to parallel, the cyclic prefix is ​​removed, and then FFT is applied back to the frequency domain. The frequency domain data transmitted in the channel is divided into two parts: pilot data known at both the transmitting and receiving ends, and useful data carrying useful information. When the signal is received, the channel state at the time of reception is generally estimated to further improve the quality of the received signal. Figure 2 The improved channel estimation algorithm within the dashed box performs channel estimation.

[0119] First, a preliminary estimation is performed using the LS algorithm. As a relatively low-complexity algorithm, it forms the basis for algorithms such as DFT and MMSE. The relationship between signal transmission and reception in the frequency domain is expressed as follows:

[0120] Y = XH + W

[0121] The LS algorithm works by using the ratio of the received pilot Y to the inserted pilot X at the transmitter to obtain the response of the pilot signal in the channel.

[0122]

[0123] The response obtained at the pilot in the frequency domain After wavelet packet denoising, the frequency domain denoising result is obtained. In terms of wavelet packet analysis, its signal denoising algorithm is basically the same as that of wavelet analysis. The difference lies in the fact that wavelet packet analysis is a more complex and flexible analysis method. Wavelet packet analysis can simultaneously decompose the upper low-frequency and high-frequency components, thus obtaining more accurate analysis results. It has the advantages of high frequency resolution, good time-frequency domain characteristics, and strong multi-resolution analysis capabilities.

[0124] The general procedure for wavelet packet denoising on a signal is as follows:

[0125] (1) Select a wavelet and determine the level to be decomposed, and then perform wavelet packet decomposition on the signal;

[0126] (2) Define the optimal wavelet packet basis and calculate the optimal tree using the entropy standard to obtain the wavelet basis that best suits the problem;

[0127] (3) Select a suitable threshold to perform threshold quantization on the wavelet packet decomposition factor;

[0128] (4) Reconstruct the wavelet packet.

[0129] At this point, the signal has undergone partial estimation and noise reduction. Next, this patent employs time-domain threshold denoising and transform-domain filter denoising to further suppress and filter out noise interference signals.

[0130] After performing wavelet packet denoising on the channel frequency domain response obtained by the LS algorithm, an N-point IDFT is performed:

[0131]

[0132] Denoising is performed on signals in the time domain. To distinguish between effective and noise values, a reasonable threshold must be set. To maximize the elimination of cyclic prefix noise while retaining as much effective signal as possible, an improved thresholding method is proposed: first, the median of the amplitude modulus of all channel responses within the cyclic prefix L is taken as the threshold; then, the average amplitude modulus of all channel responses outside the cyclic prefix is ​​taken as the threshold; finally, the sum of the two is multiplied by a suitable coefficient α to obtain the final threshold. D1, D2, and D are defined as follows:

[0133]

[0134]

[0135] D = α·(D1 + D2)

[0136] The external channel response of the CP is set to zero, while only the channel response value within the CP length is retained. Then, the internal denoising process of the CP is completed through the improved threshold processing of this patent, and the result after time-domain denoising is obtained.

[0137]

[0138] Response value after time-domain noise reduction After performing an N-point DFT transformation, the resulting frequency domain channel response value is then subjected to a DFT to obtain G(n). This domain is the transform domain. Any sequence within the transform domain is the DFT result of its corresponding sequence in the frequency domain; that is, the sequence in the transform domain is a "spectral sequence" of its corresponding sequence in the frequency domain, reflecting how quickly the channel response value changes in the frequency domain. In fading channels at sea, the channel response value changes relatively slowly, while the noise interference term changes relatively rapidly. Utilizing the characteristic that the signal in the transform domain is usually at a lower frequency, a low-pass filter can effectively filter noise while preserving the signal components in the transform domain.

[0139]

[0140] The G(n) index indicates that the signal energy is mostly distributed around n=0 and n=N-1, while noise exists throughout the entire transform domain. By setting a low-pass filter to zero for the high-frequency components of G(n) in the transform domain, we can obtain:

[0141]

[0142] Subsequently Perform IDFT to obtain the frequency domain channel response value after transform domain denoising.

[0143]

[0144] After filtering noise by combining frequency domain wavelet packet denoising techniques with appropriate threshold point selection in the time domain and transform domain denoising, the signal still needs to be equalized, converted from parallel to serial, and demodulated with the useful data. Based on the channel estimation, serial-to-parallel conversion method, and modulation method described above, the final received signal can be obtained after equalization, serial-to-parallel conversion, and demodulation.

[0145] Based on the above arguments and analysis, it is clear that using an OFDM channel estimation algorithm based on multi-domain noise reduction is feasible and has significant practical application value in maritime communications.

[0146] Example 2

[0147] According to an embodiment of the present invention, an OFDM transmission system for maritime channels is provided, characterized in that the system comprises:

[0148] The transmitter is used to send signals to the interference removal module;

[0149] The interference removal module is used to convert the signal into the time domain using IFFT and eliminate interference, and then convert it back to the frequency domain using FFT to obtain the frequency domain signal;

[0150] The channel estimation module is used to divide the frequency domain signal into useful data and pilot data, and to perform channel estimation on the pilot data.

[0151] The post-processing module is used to perform equalization, parallel-to-serial conversion, and demodulation on the channel-estimated signal and useful data, and then send the final received signal to the receiving end.

[0152] The channel estimation module performs channel estimation through the following steps:

[0153] The frequency domain signal is initially estimated using the least squares algorithm, and then the frequency domain denoising result is obtained by wavelet packet denoising.

[0154] An N-point IDFT is performed on the frequency domain denoising result. All external channel responses of the CP are set to zero, while the channel response values ​​within the CP length range are retained. The internal denoising of the CP is completed by improving the threshold processing, and the response value of the time domain denoising is obtained.

[0155] The response value of the time-domain noise reduction process is subjected to an N-point DFT transformation, and then the frequency-domain channel response value obtained by the transformation is subjected to a DFT to convert the domain to the transform domain. The noise in the transform domain is filtered by a low-pass filter to obtain the filtered data.

[0156] The filtered data is subjected to IDFT to obtain the frequency domain channel response value after transform domain denoising. Then, the channel-estimated signal is obtained through interpolation algorithm.

[0157] In one example, the initial estimate of the frequency domain signal using the least squares algorithm is as follows:

[0158] At the transmitting end, the response of the pilot signal in the channel is obtained by using the ratio of the received pilot Y to the inserted pilot X.

[0159]

[0160] in, H(k) represents the response obtained at the pilot in the frequency domain, H(k) represents the frequency response of the channel, and W(k) represents the noise during the communication process.

[0161] In one example, wavelet packet denoising includes:

[0162] Select a wavelet and determine the level to be decomposed, then perform wavelet packet decomposition on the signal;

[0163] Define the optimal wavelet packet basis, calculate the optimal tree using the entropy standard, and obtain the optimal wavelet basis;

[0164] Determine the threshold and perform threshold quantization on the wavelet packet decomposition factor;

[0165] Reconstruct the wavelet packet.

[0166] In one example, an N-point IDFT is performed using the following formula:

[0167]

[0168] in, This is the time-domain response after IDFT processing. for After wavelet packet denoising, N is the number of subcarriers, k is the frequency domain interval value, n is the time domain interval value, and j is an imaginary number.

[0169] In one example, the threshold determination for improved threshold processing includes:

[0170] Take the median value of the amplitude modulus of all channel responses within the cyclic prefix L as the threshold part:

[0171]

[0172] The threshold portion is the average amplitude of all channel responses excluding the cyclic prefix.

[0173]

[0174] The final threshold is determined by the sum of the two:

[0175] D = α·(D1 + D2)

[0176] Where D is the final threshold, D1 is the median value of the amplitude modulus of all channel responses within the cyclic prefix, D2 is the average value of the amplitude modulus of all channel responses outside the cyclic prefix, α is the multiplication coefficient, median represents the median value, and mean represents the average value.

[0177] In one example, denoising within the CP is achieved by improving thresholding:

[0178]

[0179] in, This is the response value after time-domain noise reduction processing.

[0180] In one example, a low-pass filter is used to filter transform domain noise:

[0181]

[0182] The G(n) index indicates that noise exists throughout the entire transform domain. A low-pass filter is set so that the G(n) corresponding to the high-frequency portion of the transform domain is 0:

[0183]

[0184] in, G(n) is the result after filtering noise, and G(n) is the transform domain response. c It is 1 / 2 the length of the cyclic prefix.

[0185] In one example, IDFT is performed on filtered data using the following formula:

[0186]

[0187] in, This is the frequency domain channel response value after denoising in the transform domain.

[0188] In one example, the interference removal module removes interference through the following steps:

[0189] QPSK modulation is applied to the signal at the transmitting end, and serial-to-parallel conversion is performed in complex form to obtain N parallel data streams.

[0190] A complex signal X(k) is obtained by inserting a pilot signal; the inserted pilot signal is a block pilot signal.

[0191] The signal is subjected to IFFT to obtain its time-domain form x(n);

[0192] A cyclic prefix is ​​used as the guard interval, and then the parallel-to-serial conversion is performed to send the signal into the channel;

[0193] The received signal is converted from serial to parallel, the cyclic prefix is ​​removed, and then FFT is performed to convert it back to the frequency domain to obtain the frequency domain signal.

[0194] Those skilled in the art should understand that the above description of the embodiments of the present invention is only intended to illustrate the beneficial effects of the embodiments of the present invention, and is not intended to limit the embodiments of the present invention to any of the examples given.

[0195] The various embodiments of the present invention have been described above. These descriptions are exemplary and not exhaustive, nor are they limited to the disclosed embodiments. Many modifications and variations will be apparent to those skilled in the art without departing from the scope and spirit of the described embodiments.

Claims

1. An OFDM transmission method for maritime channels, characterized in that, include: The transmitter transmits a signal, which is converted into the time domain by IFFT and interference is eliminated. Then, it is converted back into the frequency domain by FFT to obtain the frequency domain signal. The frequency domain signal is divided into useful data and pilot data, and channel estimation is performed on the pilot data. The signal after channel estimation is equalized and converted from parallel to serial, and then demodulated with the useful data to obtain the final received signal at the receiving end. The channel estimation includes: The frequency domain signal is initially estimated using the least squares algorithm, and then the frequency domain denoising result is obtained by wavelet packet denoising. An N-point IDFT is performed on the frequency domain denoising result. All external channel responses of the CP are set to zero, while the channel response values ​​within the CP length range are retained. The internal denoising of the CP is completed by improving the threshold processing, and the response value of the time domain denoising is obtained. The response value of the time-domain noise reduction process is subjected to an N-point DFT transformation, and then the frequency-domain channel response value obtained by the transformation is subjected to a DFT to convert the domain to the transform domain. The noise in the transform domain is filtered by a low-pass filter to obtain the filtered data. The filtered data is subjected to IDFT to obtain the frequency domain channel response value after transform domain denoising, and then the channel-estimated signal is obtained by interpolation algorithm; The threshold determination for the improved threshold processing includes: Take the median value of the amplitude modulus of all channel responses within the cyclic prefix L as the threshold part: The threshold portion is the average amplitude of all channel responses excluding the cyclic prefix. The final threshold is determined by the sum of the two: Where D is the final threshold, D1 is the median value of the amplitude magnitude of all channel responses within the cyclic prefix, and D2 is the average value of the amplitude magnitude of all channel responses outside the cyclic prefix. The multiplication coefficients are: median (median value) and mean (mean value). This is the time-domain response after IDFT processing. N is the number of subcarriers; The wavelet packet denoising process includes: Select a wavelet and determine the level to be decomposed, then perform wavelet packet decomposition on the signal; Define the optimal wavelet packet basis, calculate the optimal tree using the entropy standard, and obtain the optimal wavelet basis; Determine the threshold and perform threshold quantization on the wavelet packet decomposition factor; Reconstruct the wavelet packet.

2. The OFDM transmission method for maritime channels according to claim 1, wherein, The frequency domain signal is initially estimated using the least squares algorithm as follows: At the transmitting end, the response of the pilot signal in the channel is obtained by using the ratio of the received pilot Y to the inserted pilot X. in, H(k) represents the frequency response obtained at the pilot in the frequency domain, and W(k) represents the frequency response of the channel.

3. The OFDM transmission method for maritime channels according to claim 1, wherein, Perform an N-point IDFT using the following formula: in, for After wavelet packet denoising, k represents the frequency domain interval value, n represents the time domain interval value, and j represents the imaginary number.

4. The OFDM transmission method for maritime channels according to claim 1, wherein, Internal denoising of the CP is achieved by improving threshold processing: in, This is the response value after time-domain noise reduction processing.

5. The OFDM transmission method for maritime channels according to claim 1, wherein, Transform domain noise is filtered using a low-pass filter: The G(n) index indicates that noise exists throughout the entire transform domain. A low-pass filter is set so that the G(n) corresponding to the high-frequency portion of the transform domain is 0: in, This is the result after noise filtering. For the transform domain response, n c It is half the length of the cyclic prefix. This is the response value after time-domain noise reduction processing.

6. The OFDM transmission method for maritime channels according to claim 1, wherein, The IDFT is performed on the filtered data using the following formula: in, The frequency domain channel response value after transform domain denoising. This is the result after noise filtering.

7. The OFDM transmission method for maritime channels according to claim 1, wherein, The transmitted signal is converted to the time domain by IFFT and interference is eliminated. Then, it is converted back to the frequency domain by FFT to obtain the frequency domain signal, which includes: The signal from the transmitting end is QPSK modulated, and N parallel data channels are obtained by serial-to-parallel conversion in complex form. A complex signal X(k) is obtained by inserting a pilot signal; the inserted pilot signal is a block pilot signal. The signal is subjected to IFFT to obtain its time-domain form x(n); A cyclic prefix is ​​used as the guard interval, and then the parallel-to-serial conversion is performed to send the signal into the channel; The received signal is converted from serial to parallel, the cyclic prefix is ​​removed, and then FFT is performed to convert it back to the frequency domain to obtain the frequency domain signal.

8. An OFDM transmission system for maritime channels, characterized in that, include: The transmitter is used to send signals to the interference removal module; The interference removal module is used to convert the signal into the time domain via IFFT and eliminate interference, and then convert it back to the frequency domain via FFT to obtain the frequency domain signal. The channel estimation module is used to divide the frequency domain signal into useful data and pilot data, and to perform channel estimation on the pilot data. The post-processing module is used to perform equalization, parallel-to-serial conversion, and demodulation on the channel-estimated signal and the useful data, and to send the final received signal to the receiving end. The channel estimation module performs channel estimation through the following steps: The frequency domain signal is initially estimated using the least squares algorithm, and then the frequency domain denoising result is obtained by wavelet packet denoising. An N-point IDFT is performed on the frequency domain denoising result. All external channel responses of the CP are set to zero, while the channel response values ​​within the CP length range are retained. The internal denoising of the CP is completed by improving the threshold processing, and the response value of the time domain denoising is obtained. The response value of the time-domain noise reduction process is subjected to an N-point DFT transformation, and then the frequency-domain channel response value obtained by the transformation is subjected to a DFT to convert the domain to the transform domain. The noise in the transform domain is filtered by a low-pass filter to obtain the filtered data. The filtered data is subjected to IDFT to obtain the frequency domain channel response value after transform domain denoising, and then the channel-estimated signal is obtained by interpolation algorithm; The threshold determination for the improved threshold processing includes: Take the median value of the amplitude modulus of all channel responses within the cyclic prefix L as the threshold part: The threshold portion is the average amplitude of all channel responses excluding the cyclic prefix. The final threshold is determined by the sum of the two: Where D is the final threshold, D1 is the median value of the amplitude magnitude of all channel responses within the cyclic prefix, and D2 is the average value of the amplitude magnitude of all channel responses outside the cyclic prefix. The multiplication coefficients are: median (median value) and mean (mean value). This is the time-domain response after IDFT processing; The wavelet packet denoising process includes: Select a wavelet and determine the level to be decomposed, then perform wavelet packet decomposition on the signal; Define the optimal wavelet packet basis, calculate the optimal tree using the entropy standard, and obtain the optimal wavelet basis; Determine the threshold and perform threshold quantization on the wavelet packet decomposition factor; Reconstruct the wavelet packet.

Images