Short-wave channel prediction model construction system and method based on Chirp detection signal

Abstract

The invention discloses a short wave channel prediction model construction system and method based on Chirp detection signals, and belongs to the technical field of artificial intelligence. The system comprises a time window division module, a Chirp detection signal-based measurement module and a short wave channel prediction model construction module, and the time window division module divides an earth rotation period into N time windows; the measurement module based on the Chirp detection signal measures the ionization state and the optimal short wave frequency of each time window at a set place in M earth rotation periods; the short-wave channel prediction model construction module establishes a short-wave channel prediction model for each of N time windows according to the measurement data provided by the measurement module, and the short-wave channel prediction model is trained by a neural network. According to the invention, the optimal short-wave channel can be obtained for different time windows.

Description

Technical Field

[0001] The invention relates to a shortwave channel prediction model construction system and method based on Chirp detection signals, belonging to the technical field of artificial intelligence. Background Art

[0002] The Chinese invention patent application with the publication number CN103117823A discloses a shortwave channel modeling method, which includes: extracting M samples of a certain link from a database storing shortwave channel parameter samples, and obtaining multipath broadening parameters and Doppler broadening parameters in each sample; using the multipath broadening parameters and Doppler broadening parameters of each sample as column vectors, constructing a channel parameter matrix, and normalizing each vector in the matrix; defining each column of the channel parameter matrix as an array point, and calculating the cluster center of each array point using a two-dimensional clustering combination algorithm; detecting whether the number of array points covered within the neighborhood radius of each cluster center meets the clustering requirements, and if so, denormalizing the cluster center, and using the multipath broadening and Doppler broadening parameters corresponding to the denormalized cluster center as the shortwave channel model. The invention solves the problem that the model error obtained by the existing channel modeling technology is large and cannot accurately describe the link channel characteristics.

[0003] However, the shortwave channel propagated by the ionosphere is seriously affected by time, and the invention does not provide any instruction on how to select the best shortwave channel in different time periods. Summary of the invention

[0004] In order to overcome the shortcomings of the prior art, the object of the present invention is to provide a system and method for constructing a shortwave channel prediction model based on Chirp detection signals, which can establish corresponding optimal shortwave channels for different time windows.

[0005] To achieve the above-mentioned purpose, the present invention provides a shortwave channel prediction model construction system based on Chirp detection signals, which comprises: a time window division module, a measurement module and a shortwave channel prediction model construction module, wherein the time window division module divides the earth's rotation period into N time windows; the measurement module comprises an ionospheric ionization state measurement module and an optimal shortwave frequency measurement module based on Chirp detection signals, the ionospheric ionization state measurement module is used to measure the ionospheric ionization state of each time window at a set location within M earth rotation periods; the optimal shortwave frequency measurement module based on Chirp detection signals is used to measure the optimal shortwave frequency at a set location within M earth rotation periods; the shortwave channel prediction model construction module constructs a shortwave channel prediction model for each of the N time windows according to the measurement data provided by the measurement module, and the shortwave channel prediction model is trained by a neural network, and the neural network comprises an LLM model, N=2 k , k is a positive integer greater than or equal to 3, and M is a positive integer greater than or equal to 364.

[0006] To achieve the above-mentioned object of the invention, the present invention also provides a method for constructing a shortwave channel prediction model based on a Chirp detection signal, which comprises the following steps: The Earth's rotation period is divided into N time windows through the time window division module, N=2 k , k is a positive integer greater than or equal to 3; Measuring the ionospheric ionization state of each time window at a set location within M earth rotation periods by means of an ionospheric ionization state measurement module, where M is a positive integer greater than or equal to 364; The best shortwave frequency is measured at a set location within M earth rotation periods by using the best shortwave frequency measurement module based on Chirp detection signals; A shortwave channel prediction model is constructed for each of the N time windows according to the measurement data provided by the measurement module through the shortwave channel prediction model construction module. The shortwave channel prediction model is trained by a neural network.

[0007] Compared with the prior art, the shortwave channel prediction model construction system and method based on Chirp detection signal provided by the present invention has the following beneficial effects: The present invention divides the earth's rotation period into N time windows and constructs a shortwave channel prediction model for each of the N time windows, thereby obtaining the best shortwave channel for different time windows. BRIEF DESCRIPTION OF THE DRAWINGS

[0008] Figure 1 The present invention provides a block diagram of a shortwave channel prediction model building system based on Chirp detection signals.

[0009] Figure 2 It is a block diagram of the composition of the ionospheric ionization state measurement module provided by the present invention.

[0010] Figure 3 The present invention provides a block diagram of the optimal shortwave frequency measurement module based on Chirp detection signals. DETAILED DESCRIPTION

[0011] It should be noted that the preferred embodiments of the present invention will be described in detail below with reference to the accompanying drawings. The advantages and features of the present invention and methods for achieving the advantages and features will become clear with reference to the accompanying drawings and the embodiments described in detail below.

[0012] However, the present invention is not limited to the embodiments disclosed below, and can be implemented in a variety of different forms. This embodiment is only used to make the disclosure of the present invention more complete and to fully inform the scope of the invention to ordinary technicians in the technical field to which the present invention belongs. The present invention is only defined by the scope of protection claimed in the invention.

[0013] Although "first", "second", etc. are used to describe various elements, components and / or parts, these elements, components and / or parts are not limited by these terms. These terms are only used to distinguish one element, constituent element or part from other elements, constituent elements or parts. Therefore, it is obvious that within the technical spirit of the present disclosure, the first element, first element or first part mentioned below may also be the second element, second element or second part, and the terms used in this specification are only used to describe the embodiments and are not intended to limit the present disclosure.

[0014] Figure 1 The present invention provides a shortwave channel prediction model construction system based on Chirp detection signals, such as Figure 1 As shown, the shortwave channel prediction model construction system based on Chirp detection signal of the present invention comprises: a time window division module, a measurement module and a shortwave channel prediction model construction module, wherein the time window division module divides the earth rotation period into N time windows; the measurement module comprises an ionospheric ionization state measurement module and an optimal shortwave frequency measurement module based on Chirp detection signal, the ionospheric ionization state measurement module is used to measure the ionospheric ionization state of each time window at a set location within M earth rotation periods; the optimal shortwave frequency measurement module based on Chirp detection signal is used to measure the optimal shortwave frequency at a set location within M earth rotation periods; the shortwave channel prediction model construction module constructs a shortwave channel prediction model for each of the N time windows according to the measurement data provided by the measurement module, and the shortwave channel prediction model is trained by a neural network, N=2 k , k is a positive integer greater than or equal to 3.

[0015] In the present invention, if k=3, the earth's rotation period of 24 hours is divided into 8 time windows, from 0 to 3 o'clock as the first time window, 3 to 6 o'clock as the second time window, 6 to 9 o'clock as the third time window, from 9 to 12 o'clock as the fourth time window, 12 to 15 o'clock as the fifth time window, 15 to 18 o'clock as the sixth time window, 18 to 21 o'clock as the seventh time window; 21 to 24 o'clock as the eighth time window. The first to eighth shortwave channel prediction models are established for the first to eighth time windows respectively.

[0016] In the present invention, in order to make the shortwave frequency selected by the shortwave channel prediction model more accurate, the earth's rotation period can be divided into more time windows, for example, when k=4, the earth's rotation period of 24 hours is divided into 16 time windows; when k=5, the earth's rotation period of 24 hours is divided into 32 time windows, etc. A shortwave channel prediction model is established for each time window. Since the angle between the shortwave communication location and the sun is different in each time window, the ionization state of the ionosphere above the communication location is also different, so the communication effect is also different. Different shortwave channel prediction models are established for different time windows to select shortwave frequencies, which can greatly improve the quality of shortwave communication.

[0017] In the present invention, training a shortwave channel prediction model using a neural network includes the following steps: S1-1: Measure the state of the ionosphere at a set location in each time window within M Earth rotation periods; S1-2: Order ; S1-3: Get the nth time window of each Earth rotation period Ionospheric status data at the set communication location And measure the corresponding series of optimal shortwave frequencies ; S1-4: Command ; S1-5, from ionospheric state data and select the mth group of data from the optimal shortwave frequency ( , and Input to the input layer of the neural network, and output vector from the output layer of the neural network ; , is the current parameter vector of the neural network; S1-6: Set the loss function for: , The gradient is calculated according to the following formula: ; S1-7: Update the current parameter vector of the neural network according to the following formula: ; In the formula, is the learning rate, Express The gradient of S1-8: Determine the loss function Is it the minimum? If so, the nth shortwave channel prediction model is obtained, and then step S1-9 is performed; if not, the new parameter vector replaces the current parameter vector. , and return to S1-5; S1-9: Determine whether n is greater than or equal to N. If n is greater than or equal to N, obtain N shortwave channel prediction models and end; if n is less than N, n , return to step S1-3.

[0018] Figure 2 : is a block diagram of the composition of the ionospheric ionization state measurement module provided by the present invention; Figure 2 As shown, a laser source generator, an FM / AM array, an FM / AM signal generator, a splitter, a laser projector, an optical detector array, an acquisition device and a first processor, wherein the FM / AM array performs FM / AM modulation on multiple carriers of the laser source generator using multiple modulation signals generated by the FM / AM signal generator to obtain multiple FM / AM waves; the optical splitter divides the multiple FM / AM waves generated by the modulator into two paths, one for the transmitted light signal and the other for the reference light signal; the laser projector projects the transmitted light signal to the ionosphere, and the light signal reflected or scattered by the ionosphere and the reference light signal are input to the optical detector array together; the acquisition device records the amplitude modulation signal and the double frequency signal of the frequency modulation signal of each detector in the optical detector array during a unit duration τ to obtain P×2 samples, wherein P is the number of detectors in the optical detector array; wherein P is the number of detectors in the optical detector array, which is a positive integer greater than or equal to 6; the first processor acquires the ionization state of the ionosphere according to the P×2 samples.

[0019] In the first embodiment, the following matrix is ​​generated according to P×2 samples: .

[0020] The first processor includes a data preprocessing module, which calculates the concentration of the pth component contained in the ionosphere according to the following formula: : , where The carrier frequency is The amplitude modulation degree of the FM wave; The carrier frequency is The optical path length of the FM / AM wave; The receiving carrier frequency is The amplitude of the amplitude modulation signal of the FM wave; The received carrier frequency is The amplitude of the 2-fold frequency signal of the FM / AM wave; For frequency The ionosphere below contains the absorbance of the pth component, It is the frequency modulation degree of FM / AM wave; , The carrier frequency is The modulation amplitude of the FM wave; is the full width at half maximum of the absorption spectrum of the pth component contained in the ionosphere. In the present invention, yes An integer multiple of p = 1, .., P, where P is greater than or equal to 6. The components of the ionosphere include ions and molecules.

[0021] The carrier frequency signal of the FM wave is the center frequency of the mth component absorption spectrum of the ionosphere. ; 2.2 times the full width at half maximum of the absorption spectrum of the mth component contained in the ionosphere.

[0022] Figure 3 FIG. 1 is a block diagram of the optimal shortwave frequency measurement module based on the Chirp detection signal provided by the present invention. Figure 3 As shown, the optimal shortwave frequency measurement module based on the Chirp detection signal includes a transmitting device and a receiving device, and the transmitting device includes: a shortwave frequency generator, a modulator array, a Chirp detection signal generator, a power amplifier array and a transmitting antenna array, wherein the modulator array uses the Chirp detection signal generated by the Chirp detection signal generator to modulate the frequencies of multiple carriers generated by the short frequency generator to generate multiple frequency-modulated pulse electrical signals of pulses with a bandwidth B and a duration τ; the power amplifier array respectively amplifies the multiple frequency-modulated pulse electrical signals and transmits electromagnetic waves to the ionosphere through each of the transmitting antenna arrays.

[0023] The receiving device includes a receiving antenna array, a mixer array, a detector array and a second processor, wherein the receiving antenna array receives a series of multiple frequency-modulated pulse electrical signals refracted from the ionosphere; the mixer array mixes the multiple frequency-modulated pulse electrical signals refracted from the ionosphere with local oscillator telecommunications to generate multiple beat frequency signals; the detector array detects Chirp detection signals from the multiple beat frequency signals, and the second processor compares each Chirp detection signal parameter with the local Chirp reference signal. The carrier signal of the frequency-modulated pulse electrical signal with the maximum power and the most similar to the local Chirp reference signal is the optimal shortwave channel corresponding to the transmission time window.

[0024] According to another embodiment of the present invention, a small signal amplifier array is provided between the receiving antenna array and the mixer array, respectively, to perform small signal amplification on the refracted wave received by each antenna in the receiving antenna array and provide the small signal to the corresponding mixer of the mixer array.

[0025] According to another embodiment of the present invention, intermediate frequency signal amplifier arrays are provided between the mixer array and the detector array to amplify the intermediate frequency signal generated by each mixer of the mixer array and provide the signal to the corresponding detector of the detector array.

[0026] The present invention also provides a method for constructing a shortwave channel prediction model based on a Chirp detection signal, which comprises the following steps: The Earth's rotation period is divided into N time windows through the time window division module, N=2 k , k is a positive integer greater than or equal to 3; Measuring the ionospheric ionization state of each time window at a set location within M earth rotation periods by means of an ionospheric ionization state measurement module, where M is a positive integer greater than or equal to 364; The best shortwave frequency is measured at a set location within M earth rotation periods by using the best shortwave frequency measurement module based on Chirp detection signals; A shortwave channel prediction model is constructed for each of the N time windows according to the measurement data provided by the measurement module through the shortwave channel prediction model construction module. The shortwave channel prediction model is trained by a neural network.

[0027] When applying the shortwave channel prediction model provided by the present invention, the longitude difference between the preset communication location and the measurement location set when establishing the shortwave channel prediction model can be calculated first, and the angle between the communication location and the measurement location set when establishing the shortwave channel prediction model relative to the sun is calculated based on the longitude difference, and the time required for the earth to rotate by the angle is calculated based on the angle; according to the relationship between the communication location and the set measurement location, the time is added or subtracted, and the shortwave channel prediction model corresponding to the time window is selected. For example, a shortwave channel prediction model of N time windows is established by measuring the ionization state of the ionosphere at a position of 150 degrees of longitude, and the communication location is located at 180 degrees of longitude, then the communication time should be subtracted from the time taken for the earth to rotate 30 degrees, that is, 2 hours, and then the corresponding shortwave channel prediction model is used for which time window it falls into.

[0028] When applying the shortwave channel prediction model provided by the present invention, the mapping relationship between the climate at the preset communication time and the climate when the shortwave channel prediction model is established can be calculated first, and the shortwave channel prediction models of different time windows can be selected according to the mapping relationship. For example, if the climate at the preset communication time is cloudy and the shortwave channel prediction model established in the same time window is sunny, the shortwave channel prediction model of the evening time window is selected as the shortwave channel prediction model.

[0029] The above shows and describes the basic principles and main features of the present invention and the advantages of the present invention. It should be understood by those skilled in the art that the present invention is not limited to the above embodiments. The above embodiments and descriptions are only for explaining the principles of the present invention. Without departing from the spirit and scope of the present invention, the present invention may have various changes and improvements, which fall within the scope of the present invention to be protected. The scope of protection of the present invention is defined by the attached claims and their equivalents.

Claims

1. A shortwave channel prediction model construction system based on Chirp detection signal, characterized in that: include: A time window division module, a measurement module and a shortwave channel prediction model construction module, wherein the time window division module divides the earth's rotation period into N time windows; the measurement module includes an ionospheric ionization state measurement module and an optimal shortwave frequency measurement module based on a Chirp detection signal, the ionospheric ionization state measurement module is used to measure the ionospheric ionization state of each time window at a set location within M earth rotation periods; the optimal shortwave frequency measurement module based on the Chirp detection signal is used to measure the optimal shortwave frequency at a set location within M earth rotation periods; the shortwave channel prediction model construction module constructs a shortwave channel prediction model for each of the N time windows according to the measurement data provided by the measurement module, N=2 k , k is a positive integer greater than or equal to 3, M is a positive integer greater than or equal to 364, and the shortwave channel prediction model is trained by a neural network.

2. The shortwave channel prediction model construction system based on Chirp detection signal according to claim 1 is characterized in that: Training the Shengang network into a shortwave channel prediction model includes the following steps: S1-1: Measure the state of the ionosphere at a set location in each time window within M Earth rotation periods; S1-2: Order ; S1-3: Get the nth time window of each Earth rotation period Ionospheric status data at the set communication location And measure the corresponding series of optimal shortwave frequencies ; S1-4: Order ; S1-5, from ionospheric state data and select the mth group of data from the optimal shortwave frequency ( , and Input to the input layer of the neural network, and output vector from the output layer of the neural network : ，， is the current parameter vector of the neural network; S1-6: Set the loss function for: ； The gradient is calculated according to the following formula: ； S1-7: Update the current parameter vector of the neural network according to the following formula: ； In the formula, is the learning rate, Express The gradient of S1-8: Determine the loss function Is it the minimum? If so, the nth shortwave channel prediction model is obtained, and then step S1-9 is performed; if not, the new parameter vector replaces the current parameter vector. , and return to S1-5; S1-9: Determine whether n is greater than or equal to N. If n is greater than or equal to N, obtain N shortwave channel prediction models and end; if n is less than N, n , return to step S1-3.

3. The shortwave channel prediction model construction system based on Chirp detection signal according to claim 2 is characterized in that: The ionospheric ionization state measurement module includes: a laser source generator, an FM / AM array, an FM / AM signal generator, a splitter, a laser projector, an optical detector array, an acquisition device and a first processor, wherein the FM / AM array performs FM / AM modulation on multiple carriers of the laser source generator using multiple modulation signals generated by the FM / AM signal generator to obtain multiple FM / AM waves; the optical splitter divides the multiple FM / AM waves generated by the modulator into two paths, one for the transmitted light signal and the other for the reference light signal; the laser projector projects the transmitted light signal to the ionosphere, and the light signal reflected or scattered by the ionosphere and the reference light signal are input to the optical detector array together; the acquisition device records the amplitude modulation signal and the double frequency signal of the frequency modulation signal of each detector in the optical detector array during a unit duration τ to obtain P×2 samples, wherein P is the number of detectors in the optical detector array; the first processor obtains the ionization state of the ionosphere according to the P×2 samples, and P is a positive integer greater than or equal to 6.

4. The shortwave channel prediction model construction system based on Chirp detection signal according to claim 2 is characterized in that: The optimal shortwave frequency measurement module based on Chirp detection signal includes a transmitting device and a receiving device, wherein the transmitting device includes: a shortwave frequency generator, a modulator array, a Chirp detection signal generator, a power amplifier array and a transmitting antenna array, wherein the modulator array uses the Chirp detection signal generated by the Chirp detection signal generator to modulate the frequencies of multiple carriers generated by the short frequency generator to generate multiple frequency-modulated pulse electrical signals of pulses with bandwidth B and duration τ; the power amplifier array amplifies the multiple frequency-modulated pulse electrical signals and transmits electromagnetic waves to the ionosphere through each of the transmitting antenna arrays.

5. The shortwave channel prediction model construction system based on Chirp detection signal according to claim 4 is characterized in that: The receiving device includes a receiving antenna array, a mixer array, a detector array and a second processor, wherein the receiving antenna array receives a series of multiple frequency-modulated pulse electrical signals refracted from the ionosphere; the mixer array mixes the multiple frequency-modulated pulse electrical signals refracted from the ionosphere with local oscillator telecommunications to generate multiple beat frequency signals; the detector array detects Chirp detection signals from the multiple beat frequency signals, and the second processor compares each Chirp detection signal parameter with the local Chirp reference signal. The carrier signal of the frequency-modulated pulse electrical signal with the maximum power and the most similar to the local Chirp reference signal is the optimal shortwave channel corresponding to the transmission time window.

6. A method for constructing a shortwave channel prediction model based on Chirp detection signals, characterized in that: The steps include: The Earth's rotation period is divided into N time windows through the time window division module, N=2 k , k is a positive integer greater than or equal to 3; Measuring the ionospheric ionization state of each time window at a set location within M earth rotation periods by means of an ionospheric ionization state measurement module, where M is a positive integer greater than or equal to 364; The best shortwave frequency is measured at a set location within M earth rotation periods by using the best shortwave frequency measurement module based on Chirp detection signals; A shortwave channel prediction model is constructed for each of the N time windows according to the measurement data provided by the measurement module through the shortwave channel prediction model construction module. The shortwave channel prediction model is trained by a neural network.

7. The method for constructing a shortwave channel prediction model based on Chirp detection signals according to claim 6, characterized in that: Training the Shengang network into a shortwave channel prediction model includes the following steps: S1-1: Measure the state of the ionosphere at a set location in each time window within M Earth rotation periods; S1-2: Order ; S1-3: Get the nth time window of each Earth rotation period Ionospheric status data at the set communication location And measure the corresponding series of optimal shortwave frequencies ; S1-4: Order ; S1-5, from ionospheric state data and select the mth group of data from the optimal shortwave frequency ( , and Input to the input layer of the neural network, and output vector from the output layer of the neural network : ， is the current parameter vector of the neural network; S1-6: Set the loss function for: ； The gradient is calculated according to the following formula: ； S1-7: Update the current parameter vector of the neural network according to the following formula: ； In the formula, is the learning rate, Express The gradient of S1-8: Determine the loss function Is it the minimum? If so, the nth shortwave channel prediction model is obtained, and then step S1-9 is performed; if not, the new parameter vector replaces the current parameter vector. , and return to S1-5; S1-9: Determine whether n is greater than or equal to N. If n is greater than or equal to N, obtain N shortwave channel prediction models and end; if n is less than N, n , return to step S1-3.

Images