Shortwave communication frequency selection method based on historical communication record data

Abstract

The present invention discloses a short-wave communication frequency selection method based on historical communication record data. By processing the historical data accumulated during the short-wave communication process, the optimal communication frequency is predicted and selected. The frequency selection result has great relevance and accuracy, is more in line with the actual communication situation, makes up for the deficiencies of the medium- and long-term frequency prediction system in terms of real-time performance and accuracy, does not require a dedicated additional frequency prediction system, is simple, efficient, and has strong practicability.

Description

Technical Field

[0001] The present invention belongs to the field of short-wave communication, and particularly relates to a short-wave communication frequency selection method based on short-wave communication historical record data. Background Art

[0002] Traditional short-wave communication frequency prediction methods such as Chirp detection and medium- and long-term frequency prediction systems.

[0003] Chirp detection: To perform frequency prediction and selection using the Chirp detection method, an independent short-wave Chirp detection system is required. The data obtained from each real-time detection is statistically analyzed to determine the optimal communication frequency, which has high real-time performance and reliability. However, it is costly and is often applied to professional groups for ionospheric research, and is not suitable for short-wave communication units.

[0004] Medium- and long-term frequency prediction system: The ionosphere is continuously monitored through an ionospheric detection system to record the characteristics of the ionosphere, summarize the laws of solar activity and the Earth's magnetic field, and discover the relevant characteristics of the ionosphere. The medium- and long-term frequency prediction software predicts the highest available frequency, the lowest available frequency, and the optimal available frequency of the current communication link based on parameters such as the communication time, communication location (longitude and latitude), weather, sunspot number, types of transmitting and receiving antennas, and transmission power of the transceiver sites. Long-term forecasts usually use months or years as the time unit, considering the long-term change trend of the short-wave channel, and less considering the real-time situation of the short-wave channel. Therefore, it is generally used for rough selection of frequencies or to determine the available frequency bands. Summary of the Invention

[0005] Aiming at the problems existing in the existing frequency selection technologies, the purpose of the present invention is to provide a short-wave communication frequency selection method based on historical communication record data. By processing the historical data accumulated during the short-wave communication process, the optimal communication frequency is selected, which effectively guarantees the establishment of the short-wave communication link and further improves the success rate of short-wave communication link establishment.

[0006] To achieve the above purpose, the technical solution of the present invention is a short-wave communication frequency selection method based on historical communication record data, including the following steps:

[0007] Step 1: Establish a historical communication log:

[0008] In the short-wave communication system, a historical communication log file is established. The historical communication log file details the communication information of the short-wave communication system during the entire working process; the short-wave communication control system saves all historical communication logs, and the "frequency - score" information in the logs is determined during the link quality analysis. The historical communication log file is stored in the form of a text file, and each day's communication record generates an independent historical communication log file;

[0009] Step 2: Obtain the geographical location information and time information of the communication parties' sites:

[0010] The shortwave communication control system obtains the geographical location information and time information of the communication parties' sites through Beidou or GPS;

[0011] Step 3: Read the historical communication records:

[0012] Read the historical communication records in the historical communication log file in sequence;

[0013] Step 4: Screen the historical communication records according to the geographical location range and time range:

[0014] The shortwave communication control system analyzes the historical communication log according to the geographical locations of the communication parties' sites and the communication time, selects the "frequency - score" pairs that meet the geographical location range and time range from it, screens out the valid communication records, and predicts the communication frequency according to these communication records;

[0015] Step 5: Judge whether the number of the extracted historical communication records is sufficient, and perform frequency prediction and selection in different ways according to the number of the extracted historical communication records;

[0016] Step 6: Process the historical communication records, score and sort them;

[0017] After step 4 screens out the valid historical communication records, it is necessary to score the obtained historical communication record data form "frequency - score", where the highest score is 100 points and the lowest is "×", and "×" indicates that the communication of this frequency fails; since the same frequency may be used multiple times and the scores for each use are not necessarily the same, it is necessary to process the original "frequency - score" data screened out;

[0018] Step 7: Curve fitting to obtain the "frequency - score" curve:

[0019] Step 6 extracts a large number of "frequency - score" data that meet both the geographical location range and the time range from the historical communication records; uses the least - squares method for curve fitting to fit the discrete "frequency - score" historical data to obtain a smooth and continuous "frequency - score" curve, with the horizontal axis being the frequency and the vertical axis being the score. It is convenient to predict the scores of other frequencies on this curve, and sort the frequencies from high to low scores, that is, obtain the ranking result of the pros and cons of the given frequencies;

[0020] Step 8: Determine the optimal frequency according to the frequency selection conditions:

[0021] (1) Unconditionally select the optimal frequency in the full frequency band;

[0022] (2) Select the optimal frequency in the given frequency band;

[0023] (3) Select the best frequency among the given frequency points;

[0024] (4) Select the best frequency band in the full frequency band;

[0025] (5) Select one or more best frequency bands in the given frequency band.

[0026] Step 9: Eliminate interference frequencies, disabled frequencies, etc. to obtain the final frequency.

[0027] Further, the screening in step 4 includes two processes: screening historical communication records according to the geographical location range and screening historical communication records according to the time range:

[0028] (1) Screen historical communication records according to the geographical location range

[0029] Taking the geographical locations of the two communication parties' sites as the center, reasonably select a certain grid distance radius area according to the actual situation. All sites within this area are regarded as the same site. All historical communication records with the longitude and latitude of the communication location within this geographical area in the historical communication records are valid and can be used as the basis for selecting the communication frequency this time;

[0030] (2) Screen historical communication records according to the time range

[0031] Select the records that meet the communication time of T among the historical communication records that meet the geographical location range selection conditions; the historical communication records at the same time T, one hour before and after time T within a one-month range are all records that meet the time range; if the number of communication records that meet the requirements is too small, consider using the historical communication records at the same time T and one hour before and after time T within one year as the basis for predicting the communication frequency this time.

[0032] Further, step 5 uses different methods for frequency prediction and selection according to the number of historical communication records extracted. The specific methods are as follows:

[0033] (1) When there are no historical records that meet the requirements or the number is too small

[0034] After screening according to the geographical location range and time range, if there are no historical communication records that meet the requirements or the number is too small, the program flow uses frequency selection based on medium- and long-term prediction software;

[0035] (2) When the number of historical communication records that meet the requirements is sufficient

[0036] After screening according to the geographical location range and time range, when there are sufficient reference historical communication records, continue to step 6.

[0037] Further, the method for processing the selected original "frequency - score" data in step 6 is as follows:

[0038] For the case where the frequencies are the same but the scores are different, weighted processing should be performed according to the different usage times to obtain a weighted score as the final score corresponding to this frequency. In principle, the weight value for the time closer to the communication time is set higher, and the weight value for the time farther from the communication time is set lower. For the sake of simplicity in calculation, the average value of the scores is taken as the final score for this frequency.

[0039] For the frequency with a score of "×", make full use of all frequency information, assign it a score and participate in the subsequent prediction. The assigned score is selected from -10 to 10 points.

[0040] Further, the specific method for unconditionally selecting the best frequency in step 8 (1) across the entire frequency band is as follows:

[0041] First, within the entire frequency band of 1.0 - 30 MHz, select frequency points evenly at a certain frequency interval. Referring to the results of the "frequency - score" fitting curve, score the scores of each frequency point, sort the frequencies according to the scores from large to small, and select the required N frequencies with high scores as the frequency selection result across the entire frequency band.

[0042] Further, the specific method for selecting the best frequency in a given frequency band in step 8 (2) is as follows:

[0043] Within the given frequency band, select frequency points at a certain frequency interval. Referring to the results of the "frequency - score" fitting curve, predict the scores of the selected frequency points, sort the frequencies according to the scores from large to small, and take the required first N frequencies with high scores as the frequency selection result.

[0044] Further, the specific method for selecting the best frequency from given frequency points in step 8 (3) is as follows:

[0045] For all the frequencies in the given frequency points, referring to the results of the "frequency - score" fitting curve, score the scores of each frequency point, sort the frequencies according to the scores from large to small, and take the required first N frequencies with high scores as the frequency selection result.

[0046] Further, the specific method for selecting the best frequency band across the entire frequency band in step 8 (4) is as follows:

[0047] In the entire frequency band of 1.0 - 30 MHz, referring to the results of the "frequency - score" fitting curve, set a threshold score according to actual needs. The frequencies with frequency scores higher than this threshold score form one or more frequency bands, which are taken as the frequency selection result; if all the frequency scores are not higher than this threshold score, lower the threshold score to obtain a suitable - width frequency band as the frequency selection result.

[0048] Further, the specific method for step 8(5) to select one or more optimal frequency segments from the given frequency band is as follows:

[0049] Within the given frequency band, referring to the result of the "frequency - score" fitting curve, a threshold score is set. The frequencies with frequency scores higher than this threshold constitute one or more frequency segments, which are used as the frequency selection result. If all frequency scores are not higher than this threshold score, the threshold score is lowered to obtain a frequency segment with a suitable width as the frequency selection result.

[0050] The present invention has the following beneficial effects:

[0051] 1. The short - wave communication frequency selection method of the present invention based on historical communication record data, based on past actual communication records, has unique innovation compared with the medium - and long - term frequency prediction software that relies on statistical formulas obtained from past experience accumulation. It uses mathematical analysis methods to screen and process historical data, and the prediction result is more in line with the actual channel conditions, making up for the deficiencies of the medium - and long - term frequency prediction system in terms of real - time performance and accuracy, having a higher prediction accuracy, and being more practically applicable.

[0052] 2. The short - wave communication frequency selection method of the present invention based on historical communication record data, according to the historical data accumulated in the actual communication process, referring to past actual communication records, selects historical communication frequencies that meet certain requirements (geographical location range, time range, frequency - score) as the communication frequency of the current link, conducts frequency prediction and optimization, has great relevance and accuracy, does not require a dedicated additional frequency prediction system, is simple, efficient, and highly practical. Brief Description of the Drawings

[0053] Figure 1 is the frequency selection flow chart of the present invention;

[0054] Figure 2 is the schematic diagram of the historical communication log;

[0055] Figure 3 is the least - squares method frequency prediction result. Detailed Embodiment

[0056] As Figure 1 shown, a short - wave communication frequency selection method based on historical record data includes the following steps:

[0057] 1. Establish a historical communication log

[0058] Historical record data is the basis of this method. The richer the historical data is, the higher the reliability rate of successful frequency selection. The present invention uses the shortwave communication historical record data for frequency selection, that is, in the shortwave communication control system software, a communication log file is designed. The communication log file details the communication information during the entire working process of shortwave communication, including the communication contact time, the geographical information of the shortwave radio stations (latitude and longitude of both communication parties' sites), the call occurrence time, the link establishment time, the frequency for establishing the link, and the link quality analysis detection score (frequency - score) corresponding to the communication link, the state transferred to after successful link establishment, and the current system state, etc.

[0059] In each operation, the existing shortwave communication control system will save all communication logs. The "frequency - score" information in the logs is measured during link quality analysis, so these logs have important reference value and are more specific and reliable than the current medium - and long - term prediction software. At present, the log files are stored in text file form, and an independent file is generated every day for recording.

[0060] 2. Obtain the geographical location information and time information of both communication parties' sites

[0061] The shortwave communication control device obtains the geographical location information and time information of both communication parties' sites through Beidou or GPS, mainly the latitude - longitude coordinates and system time, which are important bases for frequency selection and link establishment.

[0062] 3. Read historical communication records

[0063] The first step of frequency selection based on historical communication records in the present invention is to sequentially read these log files and then screen out the valid information from them. The most important basis for data screening is the geographical location range and time range, and the magnitudes of their values directly determine the accuracy of the prediction result.

[0064] 4. Screen historical communication records according to the geographical location range and time range

[0065] Two key factors for frequency selection based on historical records are: the geographical information (latitude and longitude) of both communication parties and the communication time. Since at the same moment, the communication conditions at different communication locations vary greatly. For the same communication location, due to the change of the ionosphere over time, the communication conditions at different communication times are extremely diverse. The effectiveness of the recorded information directly affects the reliability of frequency selection based on historical communication records, and rich historical communication information provides a good guarantee for improving shortwave communication quality.

[0066] Such as Figure 2As shown in the figure, the short-wave communication control system analyzes the historical communication logs according to the geographical location information (latitude and longitude) and communication time of both communication parties, and selects "frequency-score pairs" that meet the geographical location range and time range from them. The valid communication records are screened out, and the communication frequency is predicted based on these communication records.

[0067] It mainly includes two processes: screening historical records according to the geographical location range and screening historical records according to the time range.

[0068] (1)Screen historical communication records according to the geographical location range

[0069] Taking the geographical locations (latitude and longitude) of both communication parties as the center, a certain grid distance radius is reasonably selected according to the actual situation. All stations within this area are regarded as the same station. All communication records with the longitude and latitude of the communication location in this geographical area in the historical communication records are valid and can be used as the basis for selecting the communication frequency this time.

[0070] For example, if the position of the short-wave communication main station is 120.04 degrees east longitude and 32.50 degrees north latitude, and the grid range is 5.0, then all stations within (E120.04±5.0, N32.50±5.0) can be regarded as being in the same position as the short-wave communication main station; the position of the communication slave station is 41.71 degrees east longitude and 18.33 degrees north latitude, and the grid range is 5.0, then all stations within (E41.71±5.0, N18.33±5.0) can be regarded as being in the same position as the communication slave station.

[0071] (2)Screen historical records according to the time range

[0072] Select records that meet the communication time of T from the historical communication records that meet the geographical location range selection conditions. The historical records at the same time T and within one hour before and after time T within a month are all records that meet the time range. If the number of communication records that meet the requirements is too small, the historical records at the same time T and within one hour before and after time T within one year can be considered as the basis for predicting the communication frequency this time.

[0073] 5. Judge whether the number of extracted historical communication records is sufficient, and adopt different methods for frequency prediction and selection according to the number of extracted historical communication records:

[0074] (1)There are no historical communication records that meet the requirements or the number is too small

[0075] After screening according to the geographical location and time, historical communication records that meet the conditions are obtained. If there are no historical communication records that meet the requirements or the number is too small, frequency selection based on medium- and long-term prediction software is adopted, and this method is not described in this invention.

[0076] (2) There is a sufficient number of historical communication records that meet the requirements

[0077] After filtering according to geographical location and time, there are sufficient historical communication records for reference (such as the number of historical communication records for reference is much larger than the number of frequencies required to be returned), and the subsequent process is executed.

[0078] 6. Process the historical records, score and sort them

[0079] In order to obtain a sufficient amount of historical data, the logs of the previous few months or even years need to be read. The more historical data, the more accurate the prediction result and the closer it is to the actual communication situation. When reading the log file, it is necessary to score the obtained data form "frequency - score", where the highest score is 100 points and the lowest is "×" (indicating that the communication of this frequency fails). Since the same frequency may be used multiple times and the scores for each use may not be the same, the selected original data needs to be preprocessed.

[0080] For the case where the frequencies are the same but the scores are different, weighted processing should be performed according to the different usage times to obtain a weighted score as the final score corresponding to this frequency. In principle, the weight value for the time closer to the communication time is set higher, and the weight value for the time farther from the communication time is set lower. For the sake of simplicity in calculation, the average value of the scores can be taken as the final score of this frequency.

[0081] For the frequencies with a score of "×", in order to make full use of all frequency information, a score can be assigned to it and it can participate in the subsequent prediction. There are many reasons for communication failure, such as being interfered by other signal sources or noise during communication, temporary fluctuations in the ionosphere, or this frequency is indeed not suitable for communication during this time period. The assigned score can be selected from -10 to 10 points.

[0082] 7. Curve fitting to obtain the "frequency - score" curve

[0083] As Figure 2 shown, the relationship between frequency and score is not a simple linear relationship. After selecting the geographical range and time range, a large amount of "frequency - score" data that meets both the geographical location range and the time range has been extracted from the historical communication records. Curve fitting is to use a mathematical model to fit a series of discrete data into a smooth curve, which is convenient for discovering the internal relationship between the data and analyzing the changing trend of the predicted data. The present invention mainly uses the least squares method for curve fitting to fit the discrete "frequency - score" historical data into a smooth and continuous "frequency - score" curve, with the horizontal axis being the frequency and the vertical axis being the score. It is very convenient to predict the scores of other frequencies on this curve. In the subsequent steps, the frequencies are sorted from high score to low score, and the ranking result of the advantages and disadvantages of the given frequencies is obtained.

[0084] 8. Determine the optimal frequency according to the frequency selection conditions

[0085] During shortwave communication, frequency selection generally includes the following methods:

[0086] (1) Unconditionally select the optimal frequency in the entire frequency band

[0087] First, within the entire frequency band of 1.0 - 30 MHz, select frequency points evenly at a certain frequency interval. Refer to the results of the "frequency - score" fitting curve, score the scores of each frequency point, sort the frequencies from largest to smallest according to the scores, and select the required N frequencies with high scores as the frequency selection results for the entire frequency band.

[0088] (2) Select the optimal frequency in a given frequency band

[0089] Within the given frequency band, select frequency points at a certain frequency interval. Refer to the results of the "frequency - score" fitting curve, predict the scores of the selected frequency points, sort the frequencies from largest to smallest according to the scores, and take the required first N frequencies with high scores as the frequency selection results.

[0090] (3) Select the optimal frequency among the given frequency points

[0091] For all the frequencies among the given frequency points, refer to the results of the "frequency - score" fitting curve, score the scores of each frequency point, sort the frequencies from largest to smallest according to the scores, and take the required first N frequencies with high scores as the frequency selection results.

[0092] (4) Select the optimal frequency band in the entire frequency band

[0093] In the entire frequency band of 1.0 - 30 MHz, refer to the results of the "frequency - score" fitting curve, set a threshold score according to actual needs. The frequencies with frequency scores higher than this threshold score form one or more frequency bands, which are used as the frequency selection results; if all frequency scores are not higher than this threshold score, lower the threshold score to obtain a frequency band with a suitable width as the frequency selection result.

[0094] (5) Select one or more optimal frequency bands in a given frequency band

[0095] Within the given frequency band, refer to the results of the "frequency - score" fitting curve, set a threshold score. The frequencies with frequency scores higher than this threshold score form one or more frequency bands, which are used as the frequency selection; if all frequency scores are not higher than this threshold score, lower the threshold score to obtain a frequency band with a suitable width as the frequency selection result.

[0096] 9. Eliminate interference frequencies and prohibited frequencies, etc., to obtain the final frequency

[0097] In actual communication, the following situations need to be considered when the shortwave communication system selects frequencies:

[0098] (1) Avoid co-channel interference and adjacent-channel interference. The frequency interval cannot be too narrow. The maximum bandwidth of the current shortwave communication frequency band is 3 KHz. Therefore, when selecting frequencies, the interval between two frequencies cannot be less than 3 KHz.

[0099] (2) Avoid third-order intermodulation interference. The frequencies in the frequency prediction do not meet the conditions for third-order intermodulation interference, that is, any three frequencies do not satisfy,, so as to avoid third-order intermodulation interference.

[0100] (3) Avoid conflicts with prohibited frequencies or protected frequencies. During the frequency selection process, it is necessary to eliminate the frequency points that are the same as the national maritime frequencies, emergency frequencies, etc., to avoid interfering with public service frequencies.

[0101] As described above, this is only the specific implementation manner of the present invention. Obviously, the described embodiments are only a part of the embodiments of the present invention, rather than all the embodiments. Based on the embodiments of the present invention, all other implementation manners obtained by those of ordinary skill in the art without creative efforts belong to the scope protected by the present invention. Any changes or substitutions that can be easily thought of by those skilled in the art within the technical scope disclosed by the present invention should be covered by the protection scope of the present invention.

Claims

1. A shortwave communication frequency selection method based on historical communication record data, characterized in that: The following steps are involved: Step 1: Create a historical communication log: In the shortwave communication control system, a historical communication log file is established, which records the information during the communication process in detail. The shortwave communication control system will save all historical communication logs. The "frequency-score" information in the log is determined during the link quality analysis. The historical communication log file is stored in the form of a text file, and each day's communication records generate an independent historical communication log file. Step 2: Get the geographic location and time information of both communicating sites: The shortwave communication control system obtains the geographical location information and time information of both communication sites through Beidou or GPS; Step 3: Read historical communication records: Read the records in the historical communication log file in sequence; Step 4: Filter historical communication records based on geographic location and time range: The shortwave communication control system analyzes historical communication logs based on the geographic locations of both communicating parties and the communication time, and selects "frequency-score" pairs that meet the geographic location range and time range of both communicating parties, thus screening out valid historical communication records. Step 5: Determine whether the number of extracted historical communication records is sufficient, and use different methods to perform frequency prediction and selection based on the number of extracted historical communication records; Step 6: Process historical communication records and rank them; After filtering out valid historical communication records in step 4, the acquired historical communication record data in the form of "Frequency-Score" needs to be scored, with the highest score being 100 and the lowest being "×". "×" indicates communication failure at this frequency. Because the same frequency may be used multiple times and the score may not be the same each time, the filtered raw "Frequency-Score" data needs to be processed. Step 7: Curve fitting to obtain the "frequency-score" curve; A large number of "frequency-score" data points that fit both the geographic location and time range were extracted from historical communication records. Least squares curve fitting was used to fit the discrete "frequency-score" data into a smooth, continuous "frequency-score" curve, with frequency on the horizontal axis and score on the vertical axis. The scores of other frequencies can be predicted from this curve, and the frequencies can be sorted from highest to lowest according to their scores, thus obtaining a ranking of the given frequencies. Step 8: Determine the optimal frequency based on the frequency selection conditions: (1) Unconditionally select the best frequency in the entire frequency band; (2) Select the best frequency in a given frequency band; (3) Select the best frequency among the given frequency points; (4) Select the best frequency band in the entire frequency band; (5) Select one or more optimal frequency bands from a given frequency band; Step 9: Eliminate interference frequencies and prohibited frequencies to obtain the final frequency; Among them, step 4 screening includes two processes: screening historical communication records based on geographical location range and screening historical communication records based on time range: (1) Filter historical communication records based on geographic location range Taking the geographical location of the communicating sites as the center, a certain grid distance radius area is reasonably selected based on the actual situation. All sites within this area are considered to be the same site. The longitude and latitude of the communication location in the historical communication records are all valid in this geographical area and serve as the basis for the frequency selection of this communication. (2) Filter historical communication records based on time range Select the records with communication time T from the historical communication records that meet the geographic location range selection conditions; historical communication records at the same time T and one hour before and after time T within a month are all records that meet the time range; if there are too few communication records that meet the requirements, consider using historical communication records at the same time T and one hour before and after time T within a year as the basis for frequency prediction of this communication.

2. The shortwave communication frequency selection method based on historical communication record data according to claim 1, characterized in that: Step 5 uses different methods to predict and select frequencies based on the number of extracted historical communication records. The specific methods are as follows: (1) When the historical records that meet the requirements do not exist or are too few After screening based on geographic location range and time range, historical communication records that meet the conditions are obtained. If historical communication records that meet the requirements do not exist or are too few in number, frequency selection based on medium- and long-term prediction software is used; (2) When the number of historical communication records that meet the requirements is sufficient After filtering based on geographic location range and time range, if sufficient historical communication records are obtained for reference, proceed to step 6.

3. The shortwave communication frequency selection method based on historical communication record data according to claim 1, characterized in that: Step 6: The method for processing the filtered original "frequency-score" data is: For the case where the frequency is the same but the scores are different, weighted processing should be performed according to the difference in usage time to obtain a weighted score as the final score corresponding to this frequency. In principle, the weight of the frequency closer to the communication time is set higher, and the weight of the frequency farther from the communication time is set lower. For the sake of simplicity of calculation, the average of the scores is taken as the final score of the frequency; For frequencies with a score of "×", make full use of all frequency information, assign it a score and participate in subsequent predictions. The assigned score is selected from -10 to 10 points.

4. The shortwave communication frequency selection method based on historical communication record data according to claim 1, characterized in that: The specific method of step 8 (1) unconditionally selecting the best frequency in the entire frequency band is: First, within the entire frequency band of 1.0 to 30 MHz, frequency points are evenly selected at a certain frequency interval. Referring to the "frequency-score" fitting curve results, the scores of each frequency point are scored. The frequencies are sorted from large to small according to the scores, and the required N frequencies with high scores are selected as the full-band frequency selection results.

5. The shortwave communication frequency selection method based on historical communication record data according to claim 1, characterized in that: Step 8 (2) The specific method for selecting the best frequency in a given frequency band is: Within a given frequency band, select frequency points at a certain frequency interval. Refer to the "frequency-score" fitting curve results to predict the scores of the selected frequency points. Then sort the frequencies from large to small according to the scores, and take the top N frequencies with the highest scores as the frequency selection results.

6. The method for selecting a frequency for shortwave communication based on historical communication record data according to claim 1, wherein: Step 8 (3) The specific method of selecting the best frequency among the given frequency points is: Score all frequencies in a given frequency point by referring to the "frequency-score" fitting curve result, sort the frequencies from large to small according to the score, and take the top N frequencies with high scores as the frequency selection results.

7. The shortwave communication frequency selection method based on historical communication record data according to claim 1, characterized in that: Step 8 (4) The specific method of selecting the best frequency band in the entire frequency band is: In the entire frequency band of 1.0 to 30 MHz, a threshold score is set based on actual needs, referring to the "frequency-score" fitting curve results. Frequencies with frequency scores higher than the threshold score constitute one or more frequency segments, and ultimately one or more frequency segments are selected as the frequency selection results. If all frequency scores are lower than the threshold score, the threshold score is lowered to obtain a frequency segment of appropriate width as the frequency selection result.

8. The method for selecting a frequency for shortwave communication based on historical communication record data according to claim 1, wherein: The specific method of step 8 (5) selecting one or more optimal frequency segments from the given frequency segments is: In a given frequency band, a threshold score is set with reference to the "frequency-score" fitting curve results. Frequencies with frequency scores higher than the threshold constitute one or more frequency bands, which are used as the frequency selection results. If all frequency scores are not higher than the threshold score, the threshold score is lowered to obtain a frequency band of appropriate width as the frequency selection result.

Images