Efficient anti-interference broadcast emission signal modulation technology

Abstract

The invention relates to the technical field of signal modulation, and discloses a high-efficiency anti-interference broadcast emission signal modulation technology, which comprises the following steps of: 1, extracting instantaneous amplitude, phase and time-frequency characteristics of a signal through Hilbert transform and short-time Fourier transform; step 2, based on a DeepSeek model, carrying out classification identification on 11 types of interference, and outputting an interference type probability matrix; 3, a modulation mode is dynamically selected according to the interference type, FH-PM modulation is started for third-order intermodulation interference, and the subcarrier frequency hopping interval is 25 kHz; and step 4, optimizing the modulation parameter group through a cost function. According to the high-efficiency anti-interference broadcast transmitted signal modulation technology, three-order intermodulation interference frequency points are avoided through FH-PM frequency hopping modulation (25kHz interval), interference avoidance is realized by combining DeepSeek real-time classification, the bit error rate is reduced by two orders of magnitude, dynamic adjustment of a modulation parameter theta is realized through Hilbert transform + STFT feature extraction and DeepSeek model linkage, and the modulation accuracy is improved. 85% of classification accuracy is still kept under the signal-to-noise ratio of-10dB.

Description

Technical Field

[0001] The present invention relates to the field of signal modulation technology, in particular to a high-efficiency anti-interference broadcast transmission signal modulation technology. Background Art

[0002] Signal modulation is the core basic technology in communication systems. Its essence is the process of embedding baseband information into a high-frequency carrier by changing certain characteristic parameters of the carrier signal, including amplitude modulation, frequency modulation and phase modulation.

[0003] The existing broadcast transmission signal modulation, traditional FM cannot dynamically adjust the modulation parameters, and cannot quickly identify the interference type, resulting in low switching efficiency for different interferences during the modulation process, resulting in poor anti-interference effect. Therefore, a specific high-efficiency anti-interference broadcast transmission signal modulation technology is proposed to solve the above-mentioned problems. Summary of the Invention

[0004] (1) Technical problems solved

[0005] In response to the shortcomings of the existing technology, the present invention provides a high-efficiency anti-interference broadcast transmission signal modulation technology, which has the advantages of quickly identifying interference types and high modulation efficiency. It solves the problem of existing broadcast transmission signal modulation, traditional FM cannot dynamically adjust modulation parameters, and cannot quickly identify interference types, resulting in low switching efficiency for different interferences during the modulation process, resulting in poor anti-interference effect.

[0006] (2) Technical solution

[0007] To achieve the above-mentioned purpose of quickly identifying interference types and achieving high modulation efficiency, the present invention provides the following technical solution: a high-efficiency anti-interference broadcast transmission signal modulation technology, comprising the following steps:

[0008] Step 1: Extract the instantaneous amplitude, phase and time-frequency characteristics of the signal through Hilbert transform and short-time Fourier transform;

[0009] Step 2: Classify and identify 11 types of interference based on the DeepSeek model and output the interference category probability matrix;

[0010] Step 3: Dynamically select the modulation mode based on the interference type. Enable FH-PM modulation for third-order intermodulation interference, and use a subcarrier frequency hopping interval of 25kHz.

[0011] Step 4: Through the cost function Optimize modulation parameter group

[0012] Where θ is the modulation parameter vector; BER k is the bit error rate of the kth subcarrier / time slot; PAPR kis the peak-to-average power ratio of the kth subcarrier / time slot; λ is the balancing factor; w k is the channel state weight; N is the total number of resource units, and N is a positive integer.

[0013] Preferably, the step 1 of extracting signal features by Hilbert transform and short-time Fourier transform further includes the following steps:

[0014] Step 101: Signal Preprocessing

[0015] Input signal filtering, using wavelet denoising or EMD decomposition to filter out high-frequency noise, retain the main component signal, sampling rate setting, sampling rate: ≥1MHz;

[0016] Step 102: Constructing the analytical signal

[0017] Hilbert transform formula

[0018] (Cauchy principal value integral)

[0019] Analytic signal generation

[0020] (The real part is the original signal, and the imaginary part is the Hilbert transform result);

[0021] Step 103: extract the instantaneous amplitude |A(t)=|z(t)| and the instantaneous phase φ(t)=∠z(t);

[0022] Step 104: Short-time Fourier transform (STFT) is used to analyze the time-frequency energy distribution H(f, t) and identify the frequency domain characteristics of the interference.

[0023] Step 105: Joint feature output, instantaneous feature vector: [A(t), φ(t), f(t)] and time-frequency feature matrix: |X(t k ,f m )|(k=1,2,...N t ,m=1,2,...N f );

[0024] Step 106: Input the above features into the DeepSeek model.

[0025] Preferably, the DeepSeek model architecture in step 2 includes:

[0026] Input layer adaptation

[0027] Use 1D convolutional layers to process 3D feature tensors (input channels = 3, output channels = 128), and use RoPE rotational position encoding to preserve temporal relationships;

[0028] backbone network

[0029] Load DeepSeek-R1 pre-trained weights and freeze the underlying Transformer parameters to maintain feature extraction capabilities;

[0030] Output layer

[0031] Remove the original language modeling head (LMHead) and add a new fully connected classification layer: nn.Linear(hidden_size,11), using Softmax activation to output probability distribution.

[0032] Preferably, the specific steps of performing interference identification based on the DeepSeek model in step 2 are:

[0033] Step 201: Use a lightweight CNN-LSTM hybrid network

[0034] Among them, CNN layer: 3 layers of convolution + pooling, extracting local features of time-frequency images;

[0035] LSTM layer: a two-layer bidirectional structure that captures temporal dependencies;

[0036] Step 202: The fully connected layer + Softmax outputs an 11-dimensional probability vector [p1, ..., p11], corresponding to 11 interference categories such as adjacent channel leakage and third-order intermodulation;

[0037] Step 203: Set the probability threshold T=0.7. If , it is determined to be "unknown interference" and the blind adaptation mode is started.

[0038] Preferably, the CNN-LSTM hybrid network adopts a cross entropy loss function with a temperature coefficient:

[0039]

[0040] Among them, y i is the one-hot encoding of the true label; z i is the logits value output by the fully connected layer; τ is the temperature coefficient, λ is the L2 regularization coefficient, and W is the weight matrix of the fully connected layer.

[0041] Preferably, the 11 types of interference are third-order intermodulation interference, adjacent channel leakage interference, pulse noise interference, blocking interference, spurious radiation interference, phase noise interference, harmonic interference, frequency offset interference, modulation distortion interference, intermodulation distortion interference and broadband noise interference.

[0042] Preferably, in step 3, dynamically selecting the modulation mode according to the interference type includes dynamically switching the modulation mode according to the interference type:

[0043] Third-order intermodulation interference, frequency hopping-phase hopping composite modulation (FH-PM) is enabled, and the subcarrier unit hopping interval is 25kHz;

[0044] Broadband noise interference, switch to CEEMDAN decomposition algorithm, decompose the signal into 6 IMF components and then selectively reconstruct;

[0045] Narrowband burst interference triggers space-time adaptive filtering, and adding a 4th-order delay unit improves the degree of freedom.

[0046] Preferably, the FH-PM technology core architecture

[0047] Carrier frequency hopping: The transmitter periodically switches the carrier frequency according to a pseudo-random sequence (such as an m-sequence or Gold code). The hopping interval is Th (typical value 25kHz), and the frequency set is recorded as

[0048] {f1,f2,...,fN};

[0049] Phase modulation superposition, in each frequency hopping dwell time Td (usually ≤400ms), the baseband signal m(t) is used to modulate the instantaneous phase of the carrier

[0050] s(t)=Acos[2πf k t+K p m(t)](k∈[1,N])

[0051] where K p is the phase sensitivity (rad / V), f k is the current frequency hopping point.

[0052] Preferably, the specific steps of FH-PM modulation are:

[0053] Step 301: Transmitter modulation step baseband signal processing

[0054] (1) The input signal m(t) is pre-filtered to limit the bandwidth to avoid spectrum diffusion caused by phase mutation; frequency hopping sequence is generated;

[0055] (2) The pseudo-random code generator outputs the frequency control word to drive the DDS (direct digital frequency synthesizer) to switch f k ;

[0056] (3) Real-time phase offset calculation: calculate the instantaneous phase offset Δφ(t) = Kp.m(t) based on the amplitude of m(t) and inject it into the carrier phase;

[0057] (4) RF synthesis output

[0058] The quadrature modulator mixes the phase-modulated I / Q signal with the hopping carrier to generate an FH-PM composite signal.

[0059] Step 302: Demodulation at the receiving end

[0060] (1) Frequency hopping synchronization: first capture the frequency hopping pattern through the synchronization header or autocorrelation algorithm to align the local frequency sequence with the transmitter;

[0061] (2) Coherent demodulation: At each frequency point fk, a Costas loop or PLL (phase-locked loop) is used to track the carrier phase and demodulate Δφ(t);

[0062] (3) Phase information restoration: integrating Δφ(t) yields the original signal: (4) RF synthesis output: The orthogonal modulator mixes the phase-modulated I / Q signal with the hopping carrier to generate an FH-PM composite signal.

[0063] (3) Beneficial effects

[0064] Compared with the existing technology, the present invention provides a highly efficient anti-interference broadcast transmission signal modulation technology, which has the following beneficial effects:

[0065] 1. This highly efficient, anti-interference broadcast signal modulation technology avoids third-order intermodulation interference frequencies through FH-PM frequency hopping modulation (25kHz interval), and combines DeepSeek real-time classification to achieve interference avoidance, reducing the bit error rate by two orders of magnitude.

[0066] 2. This efficient and anti-interference broadcast transmission signal modulation technology realizes dynamic adjustment of the modulation parameter θ through Hilbert transform + STFT feature extraction and linkage with the DeepSeek model, and still maintains an 85% classification accuracy at a -10dB signal-to-noise ratio. BRIEF DESCRIPTION OF THE DRAWINGS

[0067] Figure 1 Schematic diagram of the broadcast transmission signal modulation method of the present invention;

[0068] Figure 2 Schematic diagram of the method for extracting signal features through Hilbert transform and short-time Fourier transform of the present invention. DETAILED DESCRIPTION

[0069] The following will clearly and completely describe the technical solutions in the embodiments of the present invention in conjunction with the accompanying drawings. Obviously, the described embodiments are only part of the embodiments of the present invention, not all of the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by ordinary technicians in this field without making creative efforts are within the scope of protection of the present invention.

[0070] See also Figure 1-2 , an efficient anti-interference broadcast transmission signal modulation technology includes the following steps:

[0071] Step 1: Extract the instantaneous amplitude, phase and time-frequency characteristics of the signal through Hilbert transform and short-time Fourier transform;

[0072] Step 2: Classify and identify 11 types of interference based on the DeepSeek model and output the interference category probability matrix;

[0073] Step 3: Dynamically select the modulation mode based on the interference type. Enable FH-PM modulation for third-order intermodulation interference, and use a subcarrier frequency hopping interval of 25kHz.

[0074] Step 4: Through the cost function Optimize modulation parameter group

[0075] Wherein, θ is the modulation parameter vector; BERk is the bit error rate of the k-th subcarrier / time slot; PAPRk is the peak-to-average power ratio of the k-th subcarrier / time slot; λ is the balancing factor; wk is the channel state weight; N is the total number of resource units, and N is a positive integer.

[0076] Example 1

[0077] Extracting signal features through Hilbert transform and short-time Fourier transform in step 1 also includes the following steps:

[0078] Step 101: Signal Preprocessing

[0079] Input signal filtering, using wavelet denoising or EMD decomposition to filter out high-frequency noise, retain the main component signal, sampling rate setting, sampling rate: ≥1MHz;

[0080] Step 102: Constructing the analytical signal

[0081] Hilbert transform formula

[0082] (Cauchy principal value integral)

[0083] Analytic signal generation

[0084] (The real part is the original signal, and the imaginary part is the Hilbert transform result);

[0085] Step 103: extract the instantaneous amplitude |A(t)=|z(t)| and the instantaneous phase φ(t)=∠z(t);

[0086] Step 104: Short-time Fourier transform (STFT) is used to analyze the time-frequency energy distribution H(f, t) and identify the frequency domain characteristics of the interference.

[0087] Step 105: Joint feature output, instantaneous feature vector: [A(t), φ(t), f(t)] and time-frequency feature matrix: |X(t k ,fm )|(k=1,2,...N t ,m=1,2,...N f );

[0088] Step 106: Input the above features into the DeepSeek model.

[0089] Short-time Fourier transform (STFT) feature extraction specifically includes the following steps:

[0090] S1: Time-frequency analysis parameter settings

[0091]

[0092] S2: STFT calculation

[0093] Time-spectral formula

[0094]

[0095] Where W(·) is the window function

[0096] Output time-frequency matrix

[0097] Row: Frequency axis (0 to Nyquist frequency)

[0098] Column: Time axis (split by window interval)

[0099] S3: Feature Enhancement

[0100] Marginal spectrum extraction:

[0101] (Energy accumulation throughout the entire period)

[0102] Hilbert spectrum fusion:

[0103] Align the STFT amplitude |X(t,f)| with the Hilbert instantaneous frequency f(t) to generate a time-frequency energy distribution map.

[0104] In this embodiment, the Hilbert transform breaks through the global phase limitation of the Fourier transform, realizes the precise extraction of the local instantaneous phase, generates an orthogonal analytical signal, strictly satisfies the complex plane rotation mapping, and STFT fusion, thereby solving the frequency aliasing problem of the pure Hilbert transform for multi-component signals.

[0105] Example 2

[0106] The DeepSeek model architecture in step 2 includes:

[0107] Input layer adaptation

[0108] Use 1D convolutional layers to process 3D feature tensors (input channels = 3, output channels = 128), and use RoPE rotational position encoding to preserve temporal relationships;

[0109] backbone network

[0110] Load DeepSeek-R1 pre-trained weights and freeze the underlying Transformer parameters to maintain feature extraction capabilities;

[0111] Output layer

[0112] Remove the original language modeling head (LMHead) and add a new fully connected classification layer: nn.Linear(hidden_size,11), using Softmax activation to output probability distribution.

[0113] The specific steps for interference identification based on the DeepSeek model in step 2 are:

[0114] Step 201: Use a lightweight CNN-LSTM hybrid network

[0115] Among them, CNN layer: 3 layers of convolution + pooling, extracting local features of time-frequency images;

[0116] LSTM layer: a two-layer bidirectional structure that captures temporal dependencies;

[0117] Step 202: The fully connected layer + Softmax outputs an 11-dimensional probability vector [p1, ..., p11], corresponding to 11 interference categories such as adjacent channel leakage and third-order intermodulation;

[0118] Step 203: Set the probability threshold T=0.7. If , it is determined to be "unknown interference" and the blind adaptation mode is started.

[0119] CNN-LSTM hybrid network uses a cross entropy loss function with a temperature coefficient:

[0120]

[0121] Among them, yi is the one-hot encoding of the true label; zi is the logits value output by the fully connected layer; τ is the temperature coefficient, λ is the L2 regularization coefficient, and W is the weight matrix of the fully connected layer.

[0122] The 11 types of interference are third-order intermodulation interference, adjacent channel leakage interference, pulse noise interference, blocking interference, spurious radiation interference, phase noise interference, harmonic interference, frequency offset interference, modulation distortion interference, intermodulation distortion interference and broadband noise interference.

[0123] In this embodiment, intelligent identification of interference types is achieved, classification accuracy is improved, and effective adjustments can be made quickly and conveniently according to different interferences.

[0124] Example 3

[0125] In step 3, the modulation mode is dynamically selected according to the interference type, including dynamically switching the modulation mode according to the interference type:

[0126] Third-order intermodulation interference, frequency hopping-phase hopping composite modulation (FH-PM) is enabled, and the subcarrier unit hopping interval is 25kHz;

[0127] Broadband noise interference, switch to CEEMDAN decomposition algorithm, decompose the signal into 6 IMF components and then selectively reconstruct;

[0128] Narrowband burst interference triggers space-time adaptive filtering, and adding a 4th-order delay unit improves the degree of freedom.

[0129] FH-PM technical core architecture

[0130] Carrier frequency hopping: The transmitter periodically switches the carrier frequency according to a pseudo-random sequence (such as an m-sequence or Gold code). The hopping interval is Th (typical value 25kHz), and the frequency set is recorded as

[0131] {f1,f2,...,fN};

[0132] Phase modulation superposition, in each frequency hopping dwell time Td (usually ≤400ms), the baseband signal m(t) is used to modulate the instantaneous phase of the carrier

[0133] s(t)=Acos[2πf k t+K p m(t)](k∈[1,N])

[0134] where K p is the phase sensitivity (rad / V), f k is the current frequency hopping point.

[0135] Example 3

[0136] Specific steps of FH-PM modulation:

[0137] Step 301: Transmitter modulation step baseband signal processing

[0138] (1) The input signal m(t) is pre-filtered to limit the bandwidth to avoid spectrum diffusion caused by phase mutation; frequency hopping sequence is generated;

[0139] (2) The pseudo-random code generator outputs the frequency control word to drive the DDS (direct digital frequency synthesizer) to switch fk;

[0140] (3) Real-time phase offset calculation: calculate the instantaneous phase offset Δφ(t) = Kp.m(t) based on the amplitude of m(t) and inject it into the carrier phase;

[0141] (4) RF synthesis output

[0142] The quadrature modulator mixes the phase-modulated I / Q signal with the hopping carrier to generate an FH-PM composite signal.

[0143] Step 302: Demodulation at the receiving end

[0144] (1) Frequency hopping synchronization: first capture the frequency hopping pattern through the synchronization header or autocorrelation algorithm to align the local frequency sequence with the transmitter;

[0145] (2) Coherent demodulation: At each frequency point fk, a Costas loop or PLL (phase-locked loop) is used to track the carrier phase and demodulate Δφ(t);

[0146] (3) Phase information restoration: integrating Δφ(t) yields the original signal:

[0147] (4) RF synthesis output: The orthogonal modulator mixes the phase-modulated I / Q signal with the hopping carrier to generate an FH-PM composite signal.

[0148] In this embodiment, the optimal transmission system is matched according to the physical characteristics of the interference, thereby effectively improving the throughput.

[0149] In summary, this highly efficient and anti-interference broadcast transmission signal modulation technology avoids third-order intermodulation interference frequencies through FH-PM frequency hopping modulation (25kHz interval), and combines DeepSeek real-time classification to achieve interference avoidance, reducing the bit error rate by 2 orders of magnitude.

[0150] Moreover, by linking Hilbert transform + STFT feature extraction with the DeepSeek model, dynamic adjustment of the modulation parameter θ is achieved, maintaining an 85% classification accuracy at a -10dB signal-to-noise ratio.

[0151] It should be noted that, in this document, relational terms such as first and second, etc., are used only to distinguish one entity or operation from another entity or operation, and do not necessarily require or imply the existence of any such actual relationship or order between these entities or operations. Moreover, the terms "comprises," "comprising," or any other variants thereof are intended to cover non-exclusive inclusion, so that a process, method, article, or device comprising a series of elements includes not only those elements, but also other elements not explicitly listed, or elements inherent to such process, method, article, or device. In the absence of further limitations, an element defined by the phrase "comprising a ..." does not exclude the presence of other identical elements in the process, method, article, or device comprising the element.

[0152] While embodiments of the present invention have been shown and described, it will be appreciated by those skilled in the art that various changes, modifications, substitutions, and variations may be made to these embodiments without departing from the principles and spirit of the invention, and that the scope of the invention is defined by the appended claims and their equivalents.

Claims

1. Highly efficient and anti-interference broadcast transmission signal modulation technology, characterized by: The steps include: Step 1: Extract the instantaneous amplitude, phase and time-frequency characteristics of the signal through Hilbert transform and short-time Fourier transform; Step 2: Classify and identify 11 types of interference based on the DeepSeek model and output the interference category probability matrix; Step 3: Dynamically select the modulation mode based on the interference type. Enable FH-PM modulation for third-order intermodulation interference, and use a subcarrier frequency hopping interval of 25kHz. Step 4: Through the cost function Optimize modulation parameter group Where θ is the modulation parameter vector; BER k is the bit error rate of the kth subcarrier / time slot; PAPR k is the peak-to-average power ratio of the kth subcarrier / time slot; λ is the balancing factor; w k is the channel state weight; N is the total number of resource units, and N is a positive integer.

2. The high-efficiency anti-interference broadcast transmission signal modulation technology according to claim 1, characterized in that: Extracting signal features by Hilbert transform and short-time Fourier transform in step 1 also includes the following steps: Step 101: Signal Preprocessing Input signal filtering, using wavelet denoising or EMD decomposition to filter out high-frequency noise, retain the main component signal, sampling rate setting, sampling rate: ≥1MHz; Step 102: Constructing the analytical signal Hilbert transform formula Analytic signal generation (The real part is the original signal, and the imaginary part is the Hilbert transform result); Step 103: extract the instantaneous amplitude |A(t)=|z(t)| and the instantaneous phase φ(t)=∠z(t); Step 104: Short-time Fourier transform (STFT) is used to analyze the time-frequency energy distribution H(f, t) and identify the frequency domain characteristics of the interference. Step 105: Joint feature output, instantaneous feature vector: [A(t), φ(t), f(t)] and time-frequency feature matrix: |X(t k ,f m )|(k=1,2,...N t ,m=1,2,...N f ); Step 106: Input the above features into the DeepSeek model.

3. The high-efficiency anti-interference broadcast transmission signal modulation technology according to claim 1, characterized in that: The DeepSeek model architecture in step 2 includes: Input layer adaptation Use 1D convolutional layers to process 3D feature tensors (input channels = 3, output channels = 128), and use RoPE rotational position encoding to preserve temporal relationships; backbone network Load DeepSeek-R1 pre-trained weights and freeze the underlying Transformer parameters to maintain feature extraction capabilities; Output layer Remove the original language modeling head (LMHead) and add a new fully connected classification layer: nn.Linear(hidden_size,11), using Softmax activation to output probability distribution.

4. The high-efficiency anti-interference broadcast transmission signal modulation technology according to claim 1, characterized in that: The specific steps for interference identification based on the DeepSeek model in step 2 are as follows: Step 201: Use a lightweight CNN-LSTM hybrid network Among them, CNN layer: 3 layers of convolution + pooling, extracting local features of time-frequency images; LSTM layer: a two-layer bidirectional structure that captures temporal dependencies; Step 202: The fully connected layer + Softmax outputs an 11-dimensional probability vector [p1, ..., p11], corresponding to 11 interference categories such as adjacent channel leakage and third-order intermodulation; Step 203: Set the probability threshold T=0.

7. If , it is determined to be "unknown interference" and the blind adaptation mode is started.

5. The high-efficiency anti-interference broadcast transmission signal modulation technology according to claim 1, characterized in that: The CNN-LSTM hybrid network adopts a cross entropy loss function with a temperature coefficient: Among them, y i is the one-hot encoding of the true label; z i is the logits value output by the fully connected layer; τ is the temperature coefficient, λ is the L2 regularization coefficient, and W is the weight matrix of the fully connected layer.

6. The high-efficiency anti-interference broadcast transmission signal modulation technology according to claim 1, characterized in that: The 11 types of interference are third-order intermodulation interference, adjacent channel leakage interference, impulse noise interference, blocking interference, spurious radiation interference, phase noise interference, harmonic interference, frequency offset interference, modulation distortion interference, intermodulation distortion interference and broadband noise interference.

7. The high-efficiency anti-interference broadcast transmission signal modulation technology according to claim 1, characterized in that: In step 3, dynamically selecting a modulation mode according to the interference type includes dynamically switching the modulation mode according to the interference type: Third-order intermodulation interference, frequency hopping-phase hopping composite modulation (FH-PM) is enabled, and the subcarrier unit hopping interval is 25kHz; Broadband noise interference, switch to CEEMDAN decomposition algorithm, decompose the signal into 6 IMF components and then selectively reconstruct; Narrowband burst interference triggers space-time adaptive filtering, and adding a 4th-order delay unit improves the degree of freedom.

8. The high-efficiency anti-interference broadcast transmission signal modulation technology according to claim 1, characterized in that: The FH-PM technology core architecture Carrier frequency hopping: The transmitter periodically switches the carrier frequency according to a pseudo-random sequence (such as an m-sequence or Gold code). The hopping interval is Th (typical value 25kHz), and the frequency set is recorded as {f1, f2, ..., fN}; Phase modulation superposition, in each frequency hopping dwell time Td (usually ≤400ms), the baseband signal m(t) is used to modulate the instantaneous phase of the carrier s(t)=Acos[2πf k t+K p ·m(t)](k∈[1,N]) where K p is the phase sensitivity (rad / V), f k is the current frequency hopping point.

9. The high-efficiency anti-interference broadcast transmission signal modulation technology according to claim 1, characterized in that: The specific steps of the FH-PM modulation are described. Step 301: Transmitter modulation step baseband signal processing (1) The input signal m(t) is pre-filtered to limit the bandwidth to avoid spectrum diffusion caused by phase mutation; frequency hopping sequence is generated; (2) The pseudo-random code generator outputs the frequency control word to drive the DDS (direct digital frequency synthesizer) to switch f k ; (3) Real-time phase offset calculation: calculate the instantaneous phase offset Δφ(t) = Kp.m(t) based on the amplitude of m(t) and inject it into the carrier phase; (4) RF synthesis output The quadrature modulator mixes the phase-modulated I / Q signal with the hopping carrier to generate an FH-PM composite signal. Step 302: Demodulation at the receiving end (1) Frequency hopping synchronization: first capture the frequency hopping pattern through the synchronization header or autocorrelation algorithm to align the local frequency sequence with the transmitter; (2) Coherent demodulation: At each frequency point fk, a Costas loop or PLL (phase-locked loop) is used to track the carrier phase and demodulate Δφ(t); (3) Phase information restoration: integrating Δφ(t) yields the original signal: (4) RF synthesis output: The orthogonal modulator mixes the phase-modulated I / Q signal with the hopping carrier to generate an FH-PM composite signal.

Images