Underwater information transmission method adaptive to air-sea cross-domain communication gateway

Abstract

The invention relates to the technical field of wireless communication, and discloses an underwater information transmission method adaptive to an air-sea cross-domain communication gateway, which is based on a communication signal modulation identification technology of multiple modulation modes, performs accurate identification on multiple modulation modes such as OFDM (Orthogonal Frequency Division Multiplexing), MASK (Multi-Amplitude Shift Keying), MQAM (Multi-Quadrature Amplitude Modulation), MPSK (Multi-Point Shift Keying) and the like through extraction of multiple characteristic parameters such as high-order cumulant, cyclic spectrum characteristics and instantaneous parameters and the like. And efficient classification is realized in combination with a decision tree classifier. The air-sea cross-domain communication gateway can dynamically adjust a self-modulation mode according to an identification result, adapts to different underwater nodes, and improves communication reliability and transmission efficiency. The method has the advantages of high reliability, low energy consumption, high adaptability and the like, can adapt to a complex dynamic marine environment, optimizes resource utilization, prolongs the service life of underwater nodes, remarkably improves the cooperative capability and stability of air-sea cross-domain communication, and provides powerful technical support for the fields of marine exploration, resource development and the like.

Description

Technical Field

[0001] The present invention relates to the field of wireless communication technology, and in particular to an underwater information transmission method adapted to an air-sea cross-domain communication gateway. Background Art

[0002] The air-sea cross-domain communication gateway provides key support for marine resource development, marine exploration and other fields by integrating air, sea and underwater communication networks. However, underwater information transmission technology faces many challenges. Underwater communication mainly relies on acoustic signals, but its bandwidth is low, the propagation delay is large and it is susceptible to noise interference; the propagation distance of optical signals and radio signals is limited underwater, which makes it difficult to meet the needs of long-distance transmission. At the same time, the complex and dynamic marine environment, such as waves, currents and random movement of nodes, increases the reliability and stability of cross-domain network connections. The energy of underwater nodes is limited, which further requires communication technology to have the characteristics of high efficiency and low energy consumption. In addition, as an important supporting technology for the development of marine resources, the air-sea cross-domain communication system needs to achieve high effectiveness and reliability in data transmission.

[0003] Therefore, in response to the above problems, there is an urgent need for an underwater information transmission method that can take into account the characteristics of effectiveness, reliability and low energy consumption, so as to adapt to the air-sea cross-domain communication gateway, realize the efficient collaboration and stable operation of multi-media heterogeneous networks, and provide strong technical support for applications in related fields. Summary of the invention

[0004] In order to solve the technical problems of the current air-sea cross-domain communication gateway in underwater information transmission, such as modulation mode mismatch, low transmission success rate in a low signal-to-noise ratio environment, and high energy consumption, the present invention provides an underwater information transmission method adapted for the air-sea cross-domain communication gateway. The present invention identifies the modulation mode of the underwater acoustic communication signal received by the air-sea cross-domain communication gateway, and adjusts the modulation mode of the air-sea cross-domain communication gateway according to the relevant information carried by it, thereby reducing the loss caused by using a communication signal modulation mode that does not match the current underwater environment. The present invention aims to provide effective and reliable underwater information transmission for nodes such as air-sea cross-domain communication gateways and underwater sensors.

[0005] An underwater information transmission method adapted to an air-sea cross-domain communication gateway provided by the present invention comprises the following steps:

[0006] S1, receiving the underwater acoustic communication signal sent by the underwater node;

[0007] S2. Identify the modulation mode of the underwater acoustic communication signal;

[0008] S3. Adjust the modulation mode of the air-sea cross-domain communication gateway according to the identified modulation mode;

[0009] S4. Use the adjusted modulation method to communicate with the corresponding underwater node.

[0010] Preferably, the modulation mode recognition algorithm includes:

[0011] a. Signal preprocessing, including separating the modulated signal from the mixed signal and filtering and noise reduction of the modulated signal;

[0012] b. Feature extraction: by analyzing the time domain characteristics and spectrum characteristics of the signal, constructing suitable feature parameters and determining the threshold for classifier classification;

[0013] c. Classifier selection: determine the type of signal by comparing the characteristic parameters of the input signal with the threshold.

[0014] Preferably, the feature extraction includes:

[0015] a. Extract high-order cumulants to suppress Gaussian white noise;

[0016] b. Extract instantaneous parameters for signal recognition in low signal-to-noise ratio environments;

[0017] c. Extract cyclic spectrum parameters to improve noise resistance performance.

[0018] Preferably, the classifier is a decision tree classifier because it is easy to understand and implement and has good real-time performance.

[0019] Preferably, the modulation mode identification based on high-order cumulants and cyclic spectrum parameters is:

[0020] a. For MASK, MQAM, MPSK, MFSK and DSSS modulation modes, they are distinguished by second- and fourth-order statistics and autocorrelation functions;

[0021] b. For OFDM modulation, the fourth-order cumulant and cyclic spectrum features are combined for identification and extraction.

[0022] Preferably, the OFDM modulation mode identification method includes:

[0023] a. Calculate the normalized generalized fourth-order cumulant estimate for separating the multi-carrier OFDM signal from the single-carrier signal;

[0024] b. Calculate parameters based on instantaneous characteristics and classify single-carrier signals into three categories: 2ASK, 4ASK and 16QAM, BPSK and 4FSK;

[0025] c. Calculate characteristic parameters based on the cyclic spectrum and use these parameters to distinguish 4ASK and 16QAM.

[0026] Compared with the related art, the underwater information transmission method adapted to the air-sea cross-domain communication gateway provided by the present invention has the following beneficial effects:

[0027] The present invention significantly improves the reliability and adaptability of underwater communication by dynamically identifying the modulation mode of underwater acoustic signals and adjusting the modulation mode of the air-sea cross-domain communication gateway in real time; and uses multiple characteristic parameters such as high-order cumulants, cyclic spectrum characteristics and instantaneous parameters, combined with a decision tree classifier, to achieve accurate identification of multiple modulation modes such as OFDM, MASK, MQAM, etc., thereby improving the success rate of information transmission in a low signal-to-noise ratio environment;

[0028] This method has the advantages of strong anti-interference ability, high reliability, and low energy consumption. It can adapt to the complex and dynamic marine environment, optimize resource utilization and extend node life, effectively improve the effectiveness and stability of air-sea cross-domain communications, and provide reliable technical support for marine exploration, resource development, etc. BRIEF DESCRIPTION OF THE DRAWINGS

[0029] Figure 1 It is a recognition flow chart combining generalized fourth-order cumulants, instantaneous characteristics, and cyclic spectra in the present invention;

[0030] Figure 2 It is the time and frequency diagram of 2ASK modulation signal and 4ASK modulation signal in the present invention;

[0031] Figure 3 It is the time and frequency diagram of 2FSK modulation signal and 4FSK modulation signal in the present invention;

[0032] Figure 4 It is a modulation and demodulation process diagram of MPSK in the present invention;

[0033] Figure 5 It is a diagram of the modulation and demodulation process of MQAM in the present invention. DETAILED DESCRIPTION

[0034] The present invention will be further described below in conjunction with the accompanying drawings and implementation modes.

[0035] The underwater information transmission method adapted to the air-sea cross-domain communication gateway proposed in the present invention mainly includes a communication signal modulation identification technology supporting multiple modulation modes. The communication signal modulation identification technology supporting multiple modulation modes is specifically implemented in the following ways:

[0036] The basic starting point of the present invention is to target the complex and changeable characteristics of underwater acoustic channels, focusing on the research of underwater acoustic communicators that support high-speed short-distance (OFDM) and various basic modulation signals in deep-sea longitudinal active communication modes, and debugging and identifying methods based on feature extraction for signal recognition. After the air-sea cross-domain communication gateway receives the underwater acoustic communication signal sent by the underwater node, it identifies the modulation mode of the signal, switches its own communication modulation mode to the corresponding modulation mode, and uses this to communicate with the corresponding underwater node.

[0037] The modulation mode recognition algorithm proposed in the present invention generally includes three aspects, namely, signal preprocessing, feature parameter selection and extraction, and classifier selection.

[0038] Signal preprocessing: including separating the modulated signal from the mixed signal, filtering and denoising the modulated signal, etc. This link is the basis for subsequent feature extraction and classification. The third chapter of this article studies how to efficiently extract the modulated signal fragment from the mixed signal fragment and perform simple denoising, which is a link in signal preprocessing.

[0039] Feature extraction: By analyzing the time domain characteristics and spectrum characteristics of the signal, construct suitable feature parameters and determine the threshold for classifier classification. This part is the key content of the recognition algorithm. Whether the selected features are appropriate will directly affect the recognition performance of the algorithm, and it is also the focus of this chapter. Among them, the high-order cumulant feature is easy to extract and can effectively suppress Gaussian white noise, but it cannot classify some signals; the instantaneous feature parameter has low complexity and simple implementation, but the recognition effect is not good in a low signal-to-noise ratio environment; the cyclic spectrum feature parameter has good noise resistance, but the calculation complexity is high.

[0040] Selection of classifier: By comparing the characteristic parameters of the input signal with the threshold, the type of signal is determined. The decision tree is easy to understand and implement, and has good real-time performance. Support vector machines and neural networks have good recognition capabilities and can adapt, but require advance training. The present invention selects a decision tree classifier.

[0041] The modulation modes of communication signals applicable to the present invention are specifically: multi-level amplitude keying (MASK), multi-level frequency shift keying (MFSK), multi-level phase shift keying (MPSK), multi-level quadrature amplitude modulation (MQAM) and OFDM modulation.

[0042] Combination Figure 2Explain Multi-Amplitude Shift Keying (MASK). MASK signal means that the information carried by the signal is expressed by the amplitude change of the carrier, and its frequency and initial phase are fixed. The baseband signal of MASK signal can be regarded as the superposition of W-1 2MASK signals with the same carrier frequency, different amplitude and different time. The formula of MASK signal is as follows:

[0043]

[0044] In the formula, A n Indicates the amplitude of the AM signal, A n ∈{A0,A1,...,A m-1}, g(t) is the baseband pulse waveform, T s is the symbol width, f c and are the carrier frequency and initial phase of the modulated signal respectively.

[0045] Combination Figure 3 Explain Multi-level Frequency Shift Keying (MFSK). MFSK signals indicate that the information carried by the signal is expressed through the frequency state change of the carrier. The frequency of the carrier of the 2FSK signal changes between two frequencies; f1 and f2. A 2FSK signal can be expressed as the sum of the superposition of two 2ASK signals with different carrier frequencies. The expression is:

[0046] e 2FSK (t)=s1(t)cos(ω1t)+s2(t)cos(ω2t)

[0047] The carrier frequency of MFSK varies between M different frequencies. Different information is represented by different carrier frequencies. If two adjacent code elements have different frequencies, the formula for MFSK can be derived as:

[0048]

[0049] Combination Figure 4 Explain multi-level phase shift keying (MPSK). The MPSK signal indicates that the information carried by the signal is expressed by the initial phase change of the carrier, while the amplitude and frequency of the signal remain fixed. The initial phases of the carrier corresponding to the information "0" and "1" of the 2PSK signal are π and 0 respectively. The generating function of the 2PSK signal is as follows:

[0050] e 2PSK (t) = s(t)cos(ω c t)

[0051] In the formula a nTake 1 or -1, when the binary symbol of the baseband signal sent is "0", a n The phase is 0; otherwise, it is -1, and the phase is π.

[0052] The MPSK formula can be derived as follows:

[0053]

[0054] In the formula is the initial phase of the modulating signal.

[0055] Combination Figure 5 Explain multi-level quadrature amplitude modulation (MQAM). In the binary ASK system, the bandwidth utilization is 1 bit / s·Hz. If the ASK signal is transmitted using orthogonal carrier modulation technology, the bandwidth utilization can be doubled. If multi-level is combined with other technologies, the bandwidth utilization can be further improved. The technology that can accomplish this task is called quadrature amplitude modulation (QAM). It is formed by performing double-sideband suppressed carrier amplitude modulation on two signals using orthogonal carriers. Usually there are binary QAM, quaternary QAM (16QAM), octal QAM (64QAM), ... The specific modulation and demodulation process simulation diagram of MQAM is shown in the figure below. Figure 5 shown.

[0056] The modulation recognition method applicable to the present invention is specifically as follows: the present invention proposes a decision-theoretic recognition scheme based on multiple characteristic parameters, including research on parameters such as high-order cumulants, instantaneous parameters and cyclic spectrum parameters.

[0057] The high-order accumulation is explained in conjunction with Table 1. The high-order accumulation of signals under different modulation modes is also very different. The signals can be classified by extracting the high-order accumulation features of the signals. The accumulation of Gaussian noise above the second order is zero, so the recognition method using high-order accumulation as a feature has a better noise suppression effect. The high-order accumulation of a Gaussian signal with a mean of zero is also zero, so the interference of Gaussian noise can be ignored when calculating the high-order accumulation of the signal. For mixed signals of non-Gaussian signals and white noise, high-order accumulation has an excellent denoising effect. By deriving the high-order accumulation of each signal through statistical theory, it can be concluded that the high-order accumulation of a Gaussian random signal is zero. These accumulations contain the characteristics of the signal, and the types of signals can be identified and classified by these characteristics. From the method for calculating the theoretical value of the high-order accumulation, the theoretical value of the accumulation of each digital modulation signal is shown in the table, where it is assumed that the energy value of the input signal is E.

[0058] Table 1 High-order accumulation

[0059]

[0060] The instantaneous characteristic parameters are: for signals with different modulation modes, their time domain instantaneous characteristic parameters are mostly different. By extracting the time domain instantaneous parameters of the modulated signal, several signals with different modulation modes can be distinguished from each other. The recognition algorithm based on instantaneous characteristic parameters has low computational complexity, but weak anti-noise performance. In a low signal-to-noise ratio environment, the signal recognition accuracy cannot be guaranteed. Suppose the result of the Hilbert transform of the input signal x(t) is y(t). The instantaneous characteristic parameters used in the present invention are: the maximum value of the zero-centered normalized instantaneous amplitude spectrum density γ max , absolute amplitude standard deviation σ aa , frequency mean M f1 , the variance of the normalized central instantaneous frequency σ af .

[0061] The maximum value of the zero-centered normalized instantaneous amplitude spectral density γ max for:

[0062] γ max =max|FFT[a cn (i)] 2 / N s |

[0063] Where N s represents the number of samples of the input signal; a(i) represents the instantaneous amplitude value of the signal; a n (i) = a(i) / m represents the normalized value of a(i); a cn (i) = a n (i)-1 is the zero-centered amplitude of a(i); m a It represents the mean of a(i), and the calculation method of a(i) is as follows:

[0064]

[0065] In order to eliminate the influence of channel gain on characteristic parameter γ to the greatest extent max The instantaneous amplitude a(i) needs to be standardized. The frequency domain modulated signal (such as 2FSK and MFSK) is not modulated in amplitude, and its envelope curve is constant, so its a(i) = 0. It can be seen that γ max = 0. For MPSK and MASK signals, their amplitude envelopes will change. cn (i)≠0, we can deduce γ max > 0. Therefore, through γ max The parameters can infer whether the envelope of the signal is constant, thereby distinguishing the two major types of signals.

[0066] Absolute amplitude standard deviation σ aa for:

[0067]

[0068] In the formula, a t To judge the strength of the signal, a t Usually it is set to 1. c represents the number of strong signals in the input signal, a cn Definition of σ aa It reflects the absolute amplitude information of the signal. The 2ASK signal does not contain normalized amplitude information. aa The value is low; the 4ASK signal contains normalized amplitude information, σ aa Higher value.

[0069] Frequency mean M f1 for:

[0070]

[0071] In the formula, f N (i) = f(i) / E(f(i))-1, where f(i) represents the instantaneous frequency of the signal, N S Indicates the number of points of the input signal. f1 Contains the time domain frequency characteristics of the signal, which can be used to distinguish MFSK and MPSK signals.

[0072] The variance of the normalized central instantaneous frequency σ af for:

[0073]

[0074] In the formula, a t The definition of and c, and the definition of f(i) are the same as above. af will be smaller than σ of 4FSK af The value can be obtained by the instantaneous parameter σ af Complete the distinction between 2FSK signals and 4FSK signals.

[0075] The present invention calculates the cyclic spectrum of the signal by a time domain smoothing period method, and the specific implementation steps are as follows:

[0076] Step 1: Calculate the spectrum of x(n) at N points when the time length is T.

[0077] Through the Venatin theorem, the power spectrum function of the signal is derived from the autocorrelation function of the signal. The derivation process is as follows:

[0078]

[0079] In the formula, is the cyclic autocorrelation function of the signal, α is the cyclic frequency, is the cyclic spectral density function of the signal.

[0080] A rectangular window (with a center time of t and a width of T) is used to intercept the finite signal x(t). The spectrum of the x(t) signal is X(f). The spectrum of the signal after the interception of the x(t) signal is defined as X T (t,f).

[0081]

[0082] In the formula, a short-time Fourier transform is performed on the x(t) signal, which is equivalent to the signal x(t) passing through a frequency filter with a center frequency of f and a bandwidth of 1 / T.

[0083] Step 2: Calculate the spectral correlation.

[0084] X T The (t,f) spectrum is shifted to both sides by α / 2 and -α / 2 respectively, and the time-varying spectrum U of the intercepted signal is obtained. T (t,f) and V T (t,f), α is the cycle frequency.

[0085] U T (t,f)=X T (t,f+α / 2)

[0086] V T (t,f)=X T (t,f-α / 2)

[0087] To U T (t,f) and V T After multiplying (t,f), the window width is taken without performing mean smoothing on the result, and finally the cyclic spectral density function is obtained.

[0088]

[0089] Wherein T is the same as T in step 1. T * (t,f) is the time-varying spectrum V T The conjugate of (t,f).

[0090] Step 3: Spectral smoothing.

[0091] Will U T (t,f) and V T Substitute the value of (t,f) into The expression is:

[0092]

[0093] Traverse each specific cyclic frequency α, calculate its cyclic spectral density function, and after the traversal is completed, obtain the complete cyclic spectrum parameter matrix of the finite signal

[0094] The specific implementation process of the debugging mode identification based on high-order accumulation and cyclic spectrum parameters proposed by the present invention is as follows:

[0095] MASK, MQAM, MPSK, MFSK and DSSS modulation modes can be distinguished by second- and fourth-order statistics and autocorrelation functions. Since OFDM modulation has sub-Gaussian distribution characteristics, conventional high-order feature extraction methods are not effective in identifying multi-carrier and single-carrier mixed signals. To address this problem, the present invention combines fourth-order cumulants with cyclic spectrum features to identify OFDM. First, the eigenvalues ​​are defined. The generalized fourth-order cumulant is defined as:

[0096] C G42 = cum(f(s(t),f(s(t),(f(s(t)) * ,(f(s(t)) * )

[0097] Where s(t) is the baseband signal, It is a custom nonlinear transformation. Define the maximum value of the zero-centered normalized instantaneous amplitude power spectrum density r max for:

[0098] r max =max{FFT[a cn (n)]·conj{FFT[a cn (n)]}} / N

[0099] Where N is the sampling length, a cn (n) represents the value of the instantaneous envelope normalized to zero as the center, and the calculation formula is a cn (n) = a(n) / m a -1, Represents the instantaneous envelope expectation. Define the standard deviation of the instantaneous phase absolute value of the zero-centered non-weak signal segment σ ap for:

[0100]

[0101] Among them, C is the instantaneous phase of the non-weak signal, a(n) represents the instantaneous amplitude, t a is the non-weak signal decision threshold, usually taking the instantaneous envelope expectation, φ NL (n) is the nonlinear instantaneous phase. Define the generalized cyclic spectrum:

[0102]

[0103] The characteristic parameter T1 represents the peak position of the generalized cyclic spectrum, and T1 represents the number of peaks.

[0104] Combination Figure 1 Describe the modulation identification process. Consider the scenario where OFDM, MASK (M = 2, 4), BPSK, 4FSK, and 16QAM exist simultaneously. Calculate the normalized generalized fourth-order cumulant estimate to separate multi-carrier OFDM from single-carrier signals. The normalized fourth-order cumulant estimate of the OFDM signal is much larger than that of other signals, and the gap gradually increases with the increase of the mixed signal signal-to-noise ratio. Therefore, by setting the decision domain reasonably, OFDM identification and extraction can be achieved. Calculate the parameters based on instantaneous characteristics, and use the parameters to divide the single-carrier signal into three categories: 2ASK, 4ASK and 16QAM, BPSK and 4FSK. The parameters are used to distinguish 4FSK from BPSK. Calculate the characteristic parameters based on the cyclic spectrum. Use this parameter to distinguish 4ASK from 16QAM.

[0105] The air-sea cross-domain communication gateway is equipped with the modulation mode identification method proposed in the present invention, identifies the modulation mode of the received communication signal, and changes its own modulation mode to adapt to the current underwater environment, thereby improving the success rate and reliability of underwater information transmission.

[0106] The above only describes the preferred embodiments of the present invention. It should be understood that the present invention is not limited to the above specific implementation methods. Although the present invention has been disclosed as above in the preferred embodiments, it is not used to limit the present invention. Any technician familiar with the profession can make some changes or modify the technical contents disclosed above into equivalent embodiments with equivalent changes without departing from the scope of the technical solution of the present invention. However, any simple modification, equivalent replacement and improvement made to the above embodiments without departing from the content of the technical solution of the present invention, based on the technical essence of the present invention, within the spirit and principles of the present invention, still fall within the protection scope of the technical solution of the present invention.

Claims

1. An underwater information transmission method adapted to an air-sea cross-domain communication gateway, characterized in that: The following steps are involved: S1, receiving the underwater acoustic communication signal sent by the underwater node; S2. Identify the modulation mode of the underwater acoustic communication signal; S3. Adjust the modulation mode of the air-sea cross-domain communication gateway according to the identified modulation mode; S4. Use the adjusted modulation method to communicate with the corresponding underwater node.

2. The underwater information transmission method adapted to the air-sea cross-domain communication gateway according to claim 1 is characterized in that: The modulation mode recognition algorithm includes: a. Signal preprocessing, including separating the modulated signal from the mixed signal and filtering and noise reduction of the modulated signal; b. Feature extraction: by analyzing the time domain characteristics and spectrum characteristics of the signal, constructing suitable feature parameters and determining the threshold for classifier classification; c. Classifier selection: Determine the type of signal by comparing the characteristic parameters of the input signal with the threshold.

3. The underwater information transmission method adapted to the air-sea cross-domain communication gateway according to claim 2 is characterized in that: The feature extraction comprises: a. Extract high-order cumulants to suppress Gaussian white noise; b. Extract instantaneous parameters for signal recognition in low signal-to-noise ratio environments; c. Extract cyclic spectrum parameters to improve noise resistance performance.

4. The underwater information transmission method adapted to the air-sea cross-domain communication gateway according to claim 2 is characterized in that: The classifier selected is a decision tree classifier because it is easy to understand and implement and has good real-time performance.

5. The underwater information transmission method adapted to the air-sea cross-domain communication gateway as claimed in claim 3 is characterized in that: The modulation mode identification based on high-order cumulants and cyclic spectrum parameters is: a. For MASK, MQAM, MPSK, MFSK and DSSS modulation modes, they are distinguished by second- and fourth-order statistics and autocorrelation functions; b. For OFDM modulation, the fourth-order cumulant and cyclic spectrum features are combined for identification and extraction.

6. The underwater information transmission method adapted to the air-sea cross-domain communication gateway according to claim 5 is characterized in that: The OFDM modulation mode identification method includes: a. Calculate the normalized generalized fourth-order cumulant estimate for separating the multi-carrier OFDM signal from the single-carrier signal; b. Calculate parameters based on instantaneous characteristics and classify single-carrier signals into three categories: 2ASK, 4ASK and 16QAM, BPSK and 4FSK; c. Calculate characteristic parameters based on the cyclic spectrum and use these parameters to distinguish 4ASK and 16QAM.

Images