Pilot-based frequency difference beamforming reception method for underwater MIMO high-speed communication

Abstract

The application relates to an underwater acoustic MIMO high-speed communication frequency difference beam forming receiving method based on a pilot, relates to the technical field of underwater acoustic communication, and aims to solve the problem that, in the prior art, a large-array array has the angle aliasing problem in array signal processing, which further leads to the problem that a communication receiver cannot obtain accurate multipath signal DOA through a conventional array signal processing algorithm, the application takes a uniform linear array with equidistant array elements and an array element spacing greater than the half wavelength of a communication signal as a receiving array, a beam forming receiver scheme suitable for a high-speed underwater acoustic communication system is provided, the angle aliasing problem of the large-array array in array signal processing is solved, and accurate multipath signal DOA can be obtained. Moreover, the application can also improve the frequency domain signal noise ratio of a received signal, so that high-reliability underwater acoustic communication is realized.

Description

Technical Field

[0001] The present invention relates to the technical field of underwater acoustic communication, and in particular to a pilot-based underwater acoustic MIMO high-speed communication frequency difference beamforming receiving method. Background Art

[0002] Underwater acoustic communication has been widely studied and applied in underwater IoT, underwater navigation, and communication networks. High data rates have become a key requirement for future underwater communications. Orthogonal frequency division multiplexing (OFDM) combined with multiple-input, multiple-output (MIMO) technology has been proven to be a robust underwater acoustic communication technology, improving data transmission rates and reliability.

[0003] Underwater acoustic channels exhibit multipath propagation characteristics. From a spatial domain perspective, multipath signals arrive at the receiving array from different directions, and each path signal has a corresponding direction of angle (DOA). Typically, MIMO receivers process the received signals one hydrophone at a time, ignoring angular domain information. However, from a communications perspective, considering the sparsity of multipath signals in the angular domain during communication signal processing can improve system performance. Existing literature has studied angular domain representation methods for underwater acoustic channels and designed spatial domain filters for detecting underwater acoustic OFDM signals. However, the algorithms proposed in these papers are only applicable to underwater acoustic communication systems where the spacing between the receiving array elements is less than or equal to half the wavelength of the carrier frequency.

[0004] Nowadays, in order to obtain higher communication rates, communication signals usually use higher carrier frequencies and larger bandwidths. If the spacing between receiving array elements is less than half a wavelength, the array aperture will be very small, resulting in reduced angular resolution. In addition, the size of some hardware devices will also be directly larger than half the wavelength of the communication signal. Therefore, the element spacing of the high-speed underwater acoustic communication receiving array is always many times larger than the wavelength. This is accompanied by the angular aliasing phenomenon in the spatial spectrum, which makes it impossible for the communication receiver to obtain accurate multipath signal DOA through conventional array signal processing algorithms, affecting the beamforming effect. Summary of the Invention

[0005] The purpose of the present invention is to propose a pilot-based underwater acoustic MIMO high-speed communication frequency difference beamforming reception method to address the problem of angle aliasing in array signal processing of large array element spacing in the prior art, which leads to the problem that the communication receiver cannot obtain accurate multipath signal DOA through conventional array signal processing algorithms.

[0006] The technical solution adopted by the present invention to solve the above technical problems is:

[0007] The pilot-based underwater acoustic MIMO high-speed communication frequency difference beamforming receiving method comprises the following steps:

[0008] Step 1: Use the underwater acoustic MIMO communication system to receive the time domain passband signal, and discretize and perform discrete Fourier transform on the time domain passband signal to obtain a frequency domain discrete signal. The underwater acoustic MIMO communication system includes N t transmitting transducers and N r receiving hydrophones;

[0009] Step 2: Perform DOA estimation on the frequency domain discrete signal to obtain the DOA estimation value of the underwater acoustic multipath channel corresponding to the transmitting transducer

[0010] Step 3: Based on the DOA estimate, the weight vector of the MVDR beamformer is obtained. The weight vector is then multiplied by the frequency domain discrete signal and the beamforming operation is performed to obtain the frequency domain signal. The frequency domain signal is then subjected to an inverse discrete Fourier transform to obtain the direction signal in the time domain.

[0011] Step 4: Direction signal in time domain Perform Doppler estimation and compensation operations to obtain direction signals;

[0012] Step 5: Perform an OFDM demodulation operation on the directional signal in step 4 to obtain an OFDM demodulated signal;

[0013] Step 6: Perform underwater acoustic channel estimation on the OFDM demodulated signal to obtain an estimated underwater acoustic channel;

[0014] Step 7: The OFDM demodulated signal and the estimated underwater acoustic channel are input into the MIMO demodulator for MIMO signal detection, and finally the recovered data bit stream, i.e., the recovered communication signal, is obtained.

[0015] Furthermore, the receiving hydrophones in the underwater acoustic MIMO communication system adopt an equally spaced uniform linear array, and the array element spacing is greater than half the wavelength of the communication signal.

[0016] Furthermore, the propagation delay between the lth channel in the linear array reaching the vth hydrophone and the lth channel reaching the reference hydrophone is expressed as:

[0017]

[0018] Where, d is the array element spacing of ULA, c is the speed of sound, is the angle at which the lth channel reaches the receiving hydrophone array, and ν is the serial number of the receiving hydrophone.

[0019] Furthermore, the time domain passband signal is expressed as:

[0020]

[0021] in, and is the channel amplitude and channel delay from the lth channel to the reference hydrophone array element, x μ (t) is the transmitted broadband signal, w ν (t) is additive white Gaussian noise, is the angle from the lth channel to the receiving hydrophone array, i.e. the DOA of the lth channel, L μ is the number of paths that the communication signal sent by the μth transmitting transducer passes through to reach the receiving array, is the propagation delay of the lth channel in the linear array to the vth hydrophone relative to the reference hydrophone.

[0022] Furthermore, the transmitted broadband signal x μ (t) is expressed as:

[0023]

[0024] Where j is the imaginary unit, t is the time index, K is the number of OFDM subcarriers, is the modulated data symbol, f k is the OFDM subcarrier frequency.

[0025] Furthermore, the specific steps of step 2 are:

[0026] Let the pilot carrier sequence of the μth transmitting transducer be expressed as p μ , the frequency difference in the frequency difference method is Then the P-FD signal corresponding to the μth transmitting transducer is expressed as:

[0027]

[0028] in * It is represented as a conjugate operation, ⊙ is the Hadamard product, f p is the frequency carrier sequence p μ The corresponding pilot frequency corresponds to the low-frequency component in the frequency difference algorithm, Y μ (f p ) is the pilot frequency f in the μth transmitting transducer p The received signal vector at Corresponding to the high-frequency component in the frequency difference algorithm, is the preprocessing direction in the frequency difference algorithm, For preprocessing direction and frequency The corresponding receiving array steering vector performs the MVDR DOA estimation algorithm on the P-FD signal to obtain the frequency f p and preprocessing direction Frequency difference MVDR spatial spectrum at Expressed as:

[0029]

[0030] in, H is the conjugate transpose operation, -1 is the matrix inversion operation, For preprocessing direction The steering vector of the receiving array corresponding to the frequency difference Δf processed by the frequency difference algorithm, is the covariance matrix, Expressed as:

[0031]

[0032] pass The multipath signal corresponding to the μth transmitting transducer is obtained at frequency f p and preprocessing direction The rough estimated DOA set on

[0033] When the preprocessing direction and the actual DOA value θ c When consistent, the estimated DOA value when hour, For one with The monotonically increasing function of Only in When the value is 0, define a pseudo-spectral function Expressed as:

[0034]

[0035] Among them, ε is the set threshold value;

[0036] The pseudo spectrum used to estimate the DOA value is finally obtained by averaging the pseudo spectra of all pilot frequencies, that is,

[0037]

[0038] Only in There is a peak when it is equal to the true DOA value, which is the final DOA estimation result.

[0039] Furthermore, the specific steps of step three are:

[0040] use Calculate the MVDR beamforming weight vector corresponding to frequency f k and the jth estimated direction The weight vector Expressed as:

[0041]

[0042] Finally, the passband frequency domain signal Y(f k ), and the passband frequency domain signal Y(f k ) and weight vector Multiply and then perform inverse discrete Fourier transform to obtain the directional signal after beamforming

[0043] Furthermore, the passband frequency domain signal Y(f k ) is expressed as:

[0044]

[0045] Among them, W(f k ) is the frequency domain noise, The reference hydrophone is at frequency f k The frequency domain received signal at Θ μ is the set vector of multipath signal DOA,

[0046] Φ(f k ,Θ μ ) is the array popularity matrix of the receiving array, Θ μ Column vector of is the steering vector of the receiving array, Expressed as:

[0047]

[0048] Where, d is the array element spacing of ULA, c is the speed of sound, is the angle at which the lth channel reaches the receiving hydrophone array.

[0049] The beneficial effects of the present invention are:

[0050] This application uses a uniform linear array with equally spaced elements and a spacing greater than half the wavelength of the communication signal as the receiving array. It proposes a beamforming receive signal processing method suitable for high-speed underwater acoustic communication systems. This method solves the angular aliasing problem in array signal processing for arrays with large element spacing, enabling accurate multipath signal Direction of Arrival (DOA). Furthermore, this application improves the frequency domain signal-to-noise ratio (SNR) of the received signal, thereby achieving highly reliable underwater acoustic communication.

[0051] The present application solves the directional aliasing problem that occurs when estimating the DOA of multipath signals due to the fact that the element spacing of an underwater acoustic communication receiving array is much larger than half the wavelength of the communication signal. BRIEF DESCRIPTION OF THE DRAWINGS

[0052] Figure 1 A block diagram of the signal processing flow for beamforming reception;

[0053] Figure 2 This is a performance comparison chart of the beamforming receive processing algorithm of this application and the traditional MIMO receive processing algorithm in a MIMO-OFDM system with 2 transmitters and 8 receivers and an array element spacing of 2.4 times the communication signal wavelength. DETAILED DESCRIPTION

[0054] It should be noted that, unless there is any conflict, the various embodiments disclosed in this application can be combined with each other.

[0055] Specific implementation method 1: refer to Figure 1 Specifically describing this embodiment, the pilot-based underwater acoustic MIMO high-speed communication frequency difference beamforming receiving method described in this embodiment includes the following steps:

[0056] Step 1: Use the underwater acoustic MIMO communication system to receive the time domain passband signal, and discretize and perform discrete Fourier transform on the time domain passband signal to obtain a frequency domain discrete signal;

[0057] Step 2: Perform DOA estimation on the frequency domain discrete signal to obtain the DOA estimation value of the underwater acoustic multipath channel corresponding to the transmitting transducer

[0058] Step 3: Based on the DOA estimate, the weight vector of the MVDR beamformer is obtained. The weight vector is then multiplied by the frequency domain discrete signal and the beamforming operation is performed to obtain the frequency domain signal. The frequency domain signal is then subjected to an inverse discrete Fourier transform to obtain the direction signal in the time domain.

[0059] Step 4: Direction signal in time domain Perform Doppler estimation and compensation operations to obtain direction signals;

[0060] Step 5: Perform an OFDM demodulation operation on the directional signal in step 4 to obtain an OFDM demodulated signal;

[0061] Step 6: Perform underwater acoustic channel estimation on the OFDM demodulated signal to obtain an estimated underwater acoustic channel;

[0062] Step 7: The OFDM demodulated signal and the estimated underwater acoustic channel are input into the MIMO demodulator for MIMO signal detection, and finally the recovered data bit stream, i.e., the recovered communication signal, is obtained.

[0063] In order to solve the problems in the existing technology, researchers have designed sparse arrays as receiving arrays for underwater acoustic communication systems, such as coprime arrays, nested arrays, and minimum redundant arrays. However, since the spacing between the array elements of the above-mentioned sparse arrays needs to be calculated in advance and the spacing between different array elements is different, it is relatively complicated to implement when the actual equipment is deployed. Chinese patent book CN110068798B proposes a four-element receiving sparse array and a receiving signal processing method for underwater autonomous vehicles. The four-element sparse array used therein has only the spacing between elements 1 and 4 greater than half a wavelength, and the spacing between the remaining three receiving array elements is still required to be less than or equal to half a wavelength. It is not suitable for communication systems where the hardware device size is greater than half a wavelength. Therefore, how to achieve unambiguous underwater acoustic channel DOA estimation under the condition that the spacing between any array element in the receiving array is greater than half a wavelength, thereby realizing high-speed underwater acoustic communication, is the focus of the research.

[0064] Based on this, this application proposes a pilot-based underwater acoustic MIMO high-speed communication frequency difference beamforming reception method. The solution idea of ​​this application is:

[0065] For N t transmitting transducers and N r The receiving end of an underwater acoustic MIMO-OFDM high-speed communication system with a receiving hydrophone adopts an equally spaced uniform linear array (ULA), and the array element spacing is set to be larger than half the wavelength of the communication signal. Assume that the communication signal emitted by the μth transmitting transducer passes through L μ The passband signal received by the νth receiving hydrophone is expressed as:

[0066]

[0067] in and is the channel amplitude and channel delay from the lth channel to the reference hydrophone array element (the top element of the hydrophone array), x μ (t) is the transmitted broadband signal, w ν (t) is additive white Gaussian noise, is the angle from the lth channel to the receiving hydrophone array, i.e. the DOA of the lth channel, is the propagation delay of the lth channel to the νth hydrophone relative to the reference hydrophone. For ULA, The expression is:

[0068]

[0069] Where d is the array element spacing of ULA, and c is the speed of sound. In this application, the transmitted signal adopts OFDM signal, so x μ (t) is expressed as:

[0070]

[0071] Where K is the number of OFDM subcarriers, is the modulated data symbol, f k is the OFDM subcarrier frequency.

[0072] This application firstly performs the received passband signal y ν (t) with sampling frequency f s Discretization, then discrete Fourier transform (DFT) is performed on the discrete time domain signal to convert it into the frequency domain, and all the hydrophones at the frequency f k The received signal in the frequency domain is represented as a vector ( T represents the transpose operation), Y(f k ) is expressed in matrix form as:

[0073]

[0074] Where W(f k ) is the frequency domain noise, The reference hydrophone is at frequency f k The frequency domain received signal at is the set vector of multipath signal DOA, Φ(f k ,Θ μ ) is the array popularity matrix of the receiving array, whose column vector is the steering vector of the array, expressed as:

[0075]

[0076] The MIMO-OFDM system involved in this application adopts a non-overlapping orthogonal pilot scheme, and the pilot subcarriers of different transmitting ends do not overlap with each other. Let the pilot carrier sequence of the μth transmitting transducer be expressed as p μ , the frequency difference in the frequency difference method is Then the P-FD signal corresponding to the μth transmitting transducer is expressed as:

[0077]

[0078] in * It is represented as a conjugate operation, ⊙ is the Hadamard product, f p is the frequency carrier sequence p μ The corresponding pilot frequency, is the preprocessing direction in the frequency difference algorithm. Execute the MVDR DOA estimation algorithm on the P-FD signal to obtain the frequency f p and preprocessing direction Frequency difference MVDR spatial spectrum at:

[0079]

[0080] in H is the conjugate transpose operation, -1 is the matrix inversion operation, is the covariance matrix, expressed as:

[0081]

[0082] By using formula (7), we can get the multipath signal corresponding to the μth transmitting transducer at frequency f p and preprocessing direction The rough estimated DOA set on When the preprocessing direction and the actual DOA value θ c When consistent, the estimated DOA value when hour, Is a The monotonically increasing function of Only in The time value is 0. Define a pseudospectral function Expressed as:

[0083]

[0084] Where ε is a threshold value set artificially. The pseudo spectrum used to estimate the DOA value is finally obtained by averaging the pseudo spectra of all pilot frequencies, that is,

[0085]

[0086] Only in There is a peak when it is equal to the true DOA value, so the final DOA estimation result is given by The peak value is given.

[0087] The fine DOA estimate of the μth transmitting transducer is expressed as Among them J μ is the number of DOA estimates corresponding to the μth transmitting transducer. Calculate the MVDR beamforming weight vector corresponding to frequency f k and the jth estimated direction The weight vector of

[0088]

[0089] Finally, the passband frequency domain signal Y(f k ) is multiplied by the weight vector and then subjected to inverse discrete Fourier transform (IDFT) to obtain the directional signal after P-FD MVDR beamforming Its direction Spatial filtering is performed and the signal-to-noise ratio is enhanced.

[0090] After obtaining the direction signal, it performs Doppler estimation and compensation, underwater acoustic channel estimation, and MIMO signal demodulation operations to finally restore the communication signal.

[0091] Since the parallel transmission signals of the MIMO system are superimposed in the channel, the pilot-based DOA estimator proposed in this application avoids the problem of being unable to distinguish similar DOAs of signals from different transmitters due to insufficient algorithm resolution.

[0092] The MIMO-OFDM communication system described in this application uses a compressed sensing (CS) algorithm for underwater acoustic channel estimation. The CS algorithm can achieve high-resolution channel estimation results, but it also places high demands on the signal-to-noise ratio (SNR) of the received signal. The MIMO beamforming receive signal processing algorithm proposed in this application can enhance the SNR of the input signal of the CS channel estimator.

[0093] Compared with the traditional MIMO receiving and processing algorithm, the P-FD MVDR beamforming receiving and processing scheme proposed in this application greatly reduces the bit error rate (BER) of the communication system under the same signal-to-noise ratio conditions. Figure 2 shown.

[0094] It should be noted that the specific embodiments are merely explanations and illustrations of the technical solutions of the present invention and cannot be used to limit the scope of protection. Any minor changes made based on the claims and description of the present invention shall still fall within the scope of protection of the present invention.

Claims

1. A pilot-based underwater acoustic MIMO high-speed communication frequency difference beamforming receiving method, characterized in that The following steps are involved: Step 1: Use the underwater acoustic MIMO communication system to receive the time domain passband signal, and discretize and perform discrete Fourier transform on the time domain passband signal to obtain a frequency domain discrete signal. The underwater acoustic MIMO communication system includes N t transmitting transducers and N r receiving hydrophones; Step 2: Perform DOA estimation on the frequency domain discrete signal to obtain the DOA estimation value of the underwater acoustic multipath channel corresponding to the transmitting transducer; Step 3: Based on the DOA estimate, the weight vector of the MVDR beamformer is obtained. The weight vector is then multiplied by the frequency domain discrete signal and the beamforming operation is performed to obtain the frequency domain signal. The frequency domain signal is then subjected to an inverse discrete Fourier transform to obtain the direction signal in the time domain. Step 4: Direction signal in time domain Perform Doppler estimation and compensation operations to obtain direction signals; Step 5: Perform an OFDM demodulation operation on the directional signal in step 4 to obtain an OFDM demodulated signal; Step 6: Perform underwater acoustic channel estimation on the OFDM demodulated signal to obtain an estimated underwater acoustic channel; Step 7: The OFDM demodulated signal and the estimated underwater acoustic channel are input into the MIMO demodulator for MIMO signal detection, and finally the recovered data bit stream, i.e., the recovered communication signal, is obtained.

2. The pilot-based underwater acoustic MIMO high-speed communication frequency difference beamforming receiving method according to claim 1 is characterized in that The receiving hydrophones in the underwater acoustic MIMO communication system adopt an equally spaced uniform linear array, and the array element spacing is greater than half the wavelength of the communication signal.

3. The pilot-based underwater acoustic MIMO high-speed communication frequency difference beamforming receiving method according to claim 2 is characterized in that The propagation delay between the lth channel reaching the vth hydrophone and the lth channel reaching the reference hydrophone in the linear array is expressed as: Where, d is the array element spacing of ULA, c is the speed of sound, is the angle at which the lth channel reaches the receiving hydrophone array, and ν is the serial number of the receiving hydrophone.

4. The pilot-based underwater acoustic MIMO high-speed communication frequency difference beamforming receiving method according to claim 1 is characterized in that The time domain passband signal is expressed as: in, and is the channel amplitude and channel delay from the lth channel to the reference hydrophone array element, x μ (t) is the transmitted broadband signal, w ν (t) is additive white Gaussian noise, is the angle from the lth channel to the receiving hydrophone array, i.e. the DOA of the lth channel, L μ is the number of paths that the communication signal sent by the μth transmitting transducer passes through to reach the receiving array, is the relative propagation delay of the lth channel reaching the νth hydrophone in the linear array and the reference hydrophone.

5. The pilot-based underwater acoustic MIMO high-speed communication frequency difference beamforming receiving method according to claim 4 is characterized in that The transmitted broadband signal x μ (t) is expressed as: Where j is the imaginary unit, t is the time index, K is the number of OFDM subcarriers, is the modulated data symbol, f k is the OFDM subcarrier frequency.

6. The pilot-based underwater acoustic MIMO high-speed communication frequency difference beamforming receiving method according to claim 5 is characterized in that The specific steps of step 2 are: Let the pilot carrier sequence of the μth transmitting transducer be expressed as p μ , the frequency difference in the frequency difference method is Then the P-FD signal corresponding to the μth transmitting transducer is expressed as: Where * represents the conjugate operation, ⊙ is the Hadamard product, f p is the frequency carrier sequence p μ The corresponding pilot frequency corresponds to the low-frequency component in the frequency difference algorithm, Y μ (f p ) is the pilot frequency f in the μth transmitting transducer p The received signal vector at Corresponding to the high-frequency component in the frequency difference algorithm, is the preprocessing direction in the frequency difference algorithm, For preprocessing direction and frequency The corresponding receiving array steering vector performs the MVDR DOA estimation algorithm on the P-FD signal to obtain the frequency f p and preprocessing direction Frequency difference MVDR spatial spectrum at Expressed as: Where H is the conjugate transpose operation, -1 is the matrix inversion operation, For preprocessing direction The steering vector of the receiving array corresponding to the frequency difference Δf processed by the frequency difference algorithm, is the covariance matrix, Expressed as: pass The multipath signal corresponding to the μth transmitting transducer is obtained at frequency f p and preprocessing direction The rough estimated DOA set on When the preprocessing direction and the actual DOA value θ c When consistent, the estimated DOA value when hour, For one with The monotonically increasing function of Only in When the value is 0, define a pseudo-spectral function Expressed as: Among them, ε is the set threshold value; The pseudo spectrum used to estimate the DOA value is finally obtained by averaging the pseudo spectra of all pilot frequencies, that is, Only in There is a peak when it is equal to the true DOA value, which is the final DOA estimation result. is the frequency difference Δf and the preprocessing direction The steering vector of the receiving array.

7. The pilot-based underwater acoustic MIMO high-speed communication frequency difference beamforming receiving method according to claim 6 is characterized in that The specific steps of step three are: use Calculate the MVDR beamforming weight vector corresponding to frequency f k and the jth estimated direction The weight vector Expressed as: Finally, the passband frequency domain signal Y(f k ), and the passband frequency domain signal Y(f k ) and weight vector Multiply and then perform inverse discrete Fourier transform to obtain the directional signal after beamforming is the corresponding frequency f k and the jth estimated direction The array steering vector.

8. The pilot-based underwater acoustic MIMO high-speed communication frequency difference beamforming receiving method according to claim 7 is characterized in that The passband frequency domain signal Y(f k ) is expressed as: Among them, W(f k ) is the frequency domain noise, The reference hydrophone is at frequency f k The frequency domain received signal at Θ μ is the set vector of multipath signal DOA, Φ(f k ,Θ μ ) is the array popularity matrix of the receiving array, Θ μ Column vector of is the steering vector of the receiving array, Expressed as: Where, d is the array element spacing of ULA, c is the speed of sound, is the angle at which the lth channel reaches the receiving hydrophone array.

Images