Marine electromagnetic anti-interference method and system based on software radio broadband ad hoc network

Abstract

The embodiment of the present invention provides a method and system for electromagnetic anti-interference based on software radio broadband ad hoc network at sea. Since the present invention can perform feature aggregation on event environment factor variables of the maritime broadband ad hoc network status detection report and electromagnetic interference factor variables corresponding to abnormal communication status identification records, a global maritime communication factor variable corresponding to each aggregation channel of the communication status reporting event in the maritime broadband ad hoc network status detection report is obtained. In this way, feature linkage operation can be performed on the global maritime communication factor variables of different aggregation channels to obtain maritime communication linkage factor variables with more interactive features in the maritime broadband ad hoc network status detection report, so that electromagnetic anti-interference decision can be made for the maritime broadband ad hoc network status detection report based on the maritime communication linkage factor variables, thereby improving the generation accuracy and pertinence of the electromagnetic anti-interference decision plan corresponding to the maritime broadband ad hoc network status detection report.

Description

Technical Field

[0001] The present invention relates to the technical field of data processing, and in particular to a method and system for marine electromagnetic anti-interference based on software radio broadband ad hoc network. Background Art

[0002] Software radio technology is a radio communication system based on software implementation, which can simulate the functions of traditional hardware radio equipment at the software level. This technology makes the performance and functions of radio equipment more flexible and scalable, and the functions of the equipment can be changed or enhanced by simply updating the software.

[0003] Maritime communication is an important means of information exchange between ships, offshore platforms and other offshore facilities, as well as between them and land base stations. However, maritime communication faces many challenges, one of the biggest problems is electromagnetic interference. Electromagnetic interference can interfere with communication signals, resulting in a decrease in communication quality or even interruption. Therefore, how to effectively resist electromagnetic interference and ensure the stability and quality of maritime communication is an urgent problem to be solved in the current maritime communication field.

[0004] However, existing electromagnetic anti-interference strategies are usually static and cannot be dynamically adjusted according to the actual maritime communication environment and status. In addition, existing strategies often ignore the interactive influence relationship between maritime communication elements and cannot accurately reflect the overall situation of maritime communications. These factors limit the effectiveness and applicability of existing electromagnetic anti-interference strategies. Summary of the invention

[0005] In order to improve the above problems, the present invention provides a method and system for marine electromagnetic anti-interference based on software radio broadband ad hoc network.

[0006] In a first aspect, a method for marine electromagnetic anti-interference based on software radio broadband ad hoc network is provided, which is applied to a marine electromagnetic anti-interference processing system, and the method comprises:

[0007] Obtaining a state detection report of a maritime broadband self-organizing network to be optimized for electromagnetic interference resistance, and obtaining an abnormal communication state identification record corresponding to the state detection report of the maritime broadband self-organizing network;

[0008] Performing environmental factor variable mining on the maritime broadband self-organizing network status detection report to obtain event environmental factor variables of the communication status reporting event in the maritime broadband self-organizing network status detection report;

[0009] Performing electromagnetic interference factor mining on the abnormal communication state identification record to obtain electromagnetic interference factor variables of the abnormal mask text block in the abnormal communication state identification record;

[0010] Performing feature aggregation on the event environment factor variables and the electromagnetic interference factor variables to obtain a global maritime communication factor variable corresponding to each aggregation channel of the communication status reporting event;

[0011] Performing feature linkage operation on the global maritime communication element variables of different aggregation channels to obtain maritime communication linkage element variables corresponding to the maritime broadband ad hoc network status detection report;

[0012] Based on the maritime communication linkage factor variables, an electromagnetic anti-interference decision analysis is performed on the maritime broadband self-organizing network status detection report to generate an electromagnetic anti-interference decision plan corresponding to the maritime broadband self-organizing network status detection report.

[0013] In some schemes, the feature aggregation of the event environment factor variable and the electromagnetic interference factor variable to obtain the global maritime communication factor variable corresponding to each aggregation channel of the communication status reporting event includes:

[0014] Performing involved feature aggregation on the event environment involved factor variables in the event environment factor variables and the electromagnetic interference factor variables to obtain an involved factor variable relationship network between the abnormal communication state identification record and the maritime broadband ad hoc network state detection report;

[0015] Based on the involved factor variable relationship network and the phase event environment factor variables in the event environment factor variables, the global maritime communication factor variables corresponding to each aggregation channel of the communication status reporting event are determined.

[0016] In some schemes, determining the global maritime communication element variable corresponding to each aggregation channel of the communication status reporting event based on the involved element variable relationship network and the stage event environment element variable in the event environment element variable includes:

[0017] Performing knowledge mapping on the involved factor variable relationship network to obtain an involved factor mapping knowledge relationship network corresponding to the involved factor variable relationship network;

[0018] Performing knowledge mapping on the stage event environment element variables in the event environment element variables to obtain the stage event environment element mapping knowledge corresponding to the stage event environment element variables;

[0019] Based on the involved element mapping knowledge relationship network and the stage event environment element mapping knowledge, the global maritime communication element variable corresponding to each aggregation channel of the communication status reporting event is determined.

[0020] In some solutions, performing knowledge mapping on the involved factor variable relationship network to obtain an involved factor mapping knowledge relationship network corresponding to the involved factor variable relationship network includes:

[0021] Acquiring text fine-grained information of the abnormal communication state identification record;

[0022] Based on the text fine-grained information, the involved factor variable relationship network is disassembled to obtain the involved factor variable relationship subnetwork corresponding to the involved factor variable relationship network;

[0023] Performing knowledge mapping on the involved factor variable relationship subnet to obtain an involved factor mapping knowledge relationship subnet corresponding to the involved factor variable relationship subnet;

[0024] Based on the involved element mapping knowledge relationship sub-network, the involved element mapping knowledge relationship network corresponding to the involved element variable relationship network is determined.

[0025] In some schemes, determining the global maritime communication element variable corresponding to each aggregation channel of the communication status reporting event based on the involved element mapping knowledge relationship network and the stage event environment element mapping knowledge includes:

[0026] Obtaining event description keywords corresponding to the event environment element variables, and obtaining event description summary word vectors corresponding to the involved element variable relationship network;

[0027] The knowledge relationship network of the involved element mapping, the stage event environment element mapping knowledge, the event description keywords and the event description summary word vector are linked to obtain the global maritime communication element variables corresponding to each aggregation channel of the communication status reporting event.

[0028] In some schemes, the environmental factor variable mining of the maritime broadband self-organizing network status detection report to obtain the event environmental factor variable of the communication status reporting event in the maritime broadband self-organizing network status detection report includes:

[0029] Performing environmental factor variable mining on the maritime broadband ad hoc network status detection report to obtain event type factor knowledge variables, event distribution factor knowledge variables, and event impact factor knowledge variables of the communication status reporting event in the maritime broadband ad hoc network status detection report;

[0030] Based on the event type element knowledge variable, the event distribution element knowledge variable and the event impact element knowledge variable, an event environment element variable of the communication status reporting event in the maritime broadband ad hoc network status detection report is generated.

[0031] In some schemes, the generating of the event environment element variable of the communication status reporting event in the maritime broadband ad hoc network status detection report based on the event type element knowledge variable, the event distribution element knowledge variable and the event impact element knowledge variable includes:

[0032] Aggregating information features of the event type factor knowledge variable, the event distribution factor knowledge variable and the event impact factor knowledge variable to obtain original event environment factor variables of each communication status reporting event in the maritime broadband ad hoc network status detection report;

[0033] The element variable splicing operation is performed on the original event environment element variables corresponding to the different communication status reporting events to generate the event environment element variables of the communication status reporting event in the maritime broadband ad hoc network status detection report.

[0034] In some schemes, the performing of feature linkage operation on the global maritime communication element variables of different aggregated channels to obtain the maritime communication linkage element variables corresponding to the maritime broadband ad hoc network status detection report includes:

[0035] Obtaining an associated maritime communication element variable corresponding to the global maritime communication element variable;

[0036] Based on the associated maritime communication element variables, feature linkage operation is performed on the global maritime communication element variables of different aggregation channels to obtain maritime communication linkage element variables corresponding to the maritime broadband ad hoc network status detection report.

[0037] In some schemes, the performing a feature linkage operation on the global maritime communication element variables of different aggregated channels to obtain the maritime communication linkage element variables corresponding to the maritime broadband self-organizing network status detection report includes: calling a semantic feature processing branch in a natural language processing network to perform a feature linkage operation on the global maritime communication element variables of different aggregated channels to obtain the maritime communication linkage element variables corresponding to the maritime broadband self-organizing network status detection report;

[0038] The electromagnetic anti-interference decision analysis is performed on the maritime broadband self-organizing network status detection report based on the maritime communication linkage element variables to generate an electromagnetic anti-interference decision plan corresponding to the maritime broadband self-organizing network status detection report, including: calling the semantic feature processing branch in the natural language processing network, and performing electromagnetic anti-interference decision analysis on the maritime broadband self-organizing network status detection report based on the maritime communication linkage element variables to generate an electromagnetic anti-interference decision plan corresponding to the maritime broadband self-organizing network status detection report.

[0039] In some schemes, the calling of the semantic feature processing branch in the natural language processing network, based on the maritime communication linkage element variable, performs electromagnetic anti-interference decision analysis on the maritime broadband self-organizing network status detection report to generate an electromagnetic anti-interference decision plan corresponding to the maritime broadband self-organizing network status detection report, including:

[0040] Calling the semantic feature processing branch in the natural language processing network to identify the target electromagnetic interference characterization vector for the marine broadband ad hoc network status detection report based on the marine communication linkage element variable;

[0041] Based on the target electromagnetic interference characterization vector, an electromagnetic anti-interference simulation is performed on the abnormal communication status identification record to generate an electromagnetic anti-interference decision plan corresponding to the marine broadband ad hoc network status detection report.

[0042] In some schemes, the event environment factor variable is an AI knowledge factor variable obtained by calling a natural language processing network to perform environment factor variable mining on the marine broadband self-organizing network status detection report; before performing environment factor variable mining on the marine broadband self-organizing network status detection report to obtain the event environment factor variable of the communication status reporting event in the marine broadband self-organizing network status detection report, the method further includes:

[0043] Obtaining a network debugging example, wherein the network debugging example includes a maritime broadband self-organizing network state detection report learning example and an abnormal communication state identification record learning example corresponding to the maritime broadband self-organizing network state detection report learning example;

[0044] Calling the element variable processing branch of the natural language processing network to be debugged, based on the maritime broadband self-organizing network state detection report learning example and the abnormal communication state identification record learning example, obtaining the sample global maritime communication element variable corresponding to each aggregation channel of the communication state reporting event in the maritime broadband self-organizing network state detection report learning example;

[0045] Calling the semantic feature processing branch of the natural language processing network to be debugged, based on the sample global maritime communication element variable, performing a discrimination operation on the abnormal communication state identification record learning example, and obtaining an electromagnetic interference characterization vector prediction result corresponding to the abnormal communication state identification record learning example;

[0046] Based on the electromagnetic interference characterization vector in the abnormal communication state identification record learning example and the electromagnetic interference characterization vector prediction result, the natural language processing network to be debugged is debugged to obtain the natural language processing network.

[0047] In some schemes, before calling the semantic feature processing branch of the natural language processing network to be debugged, performing a discrimination operation on the abnormal communication state identification record learning example based on the sample global maritime communication element variable, and obtaining the electromagnetic interference characterization vector prediction result corresponding to the abnormal communication state identification record learning example, the method further includes:

[0048] Obtaining an associated state detection report learning example corresponding to the maritime broadband ad hoc network state detection report learning example;

[0049] Calling the element variable processing branch of the natural language processing network to be debugged, based on the association state detection report learning example and the abnormal communication state identification record learning example, obtaining the global maritime communication element variable association learning example corresponding to each aggregation channel of the communication state reporting event in the maritime broadband ad hoc network state detection report learning example;

[0050] The calling of the semantic feature processing branch of the natural language processing network to be debugged, based on the sample global maritime communication element variable, performs a discrimination operation on the abnormal communication state identification record learning example, and obtains an electromagnetic interference characterization vector prediction result corresponding to the abnormal communication state identification record learning example, including:

[0051] The semantic feature processing branch of the natural language processing network to be debugged is called, and based on the sample global maritime communication element variable and the global maritime communication element variable association learning example, a discrimination operation is performed on the abnormal communication state identification record learning example to obtain an electromagnetic interference characterization vector prediction result corresponding to the abnormal communication state identification record learning example.

[0052] In a second aspect, a marine electromagnetic anti-interference processing system is provided, comprising a processor and a memory communicating with each other, wherein the processor is used to retrieve a computer program from the memory and implement the method described in the first aspect by running the computer program.

[0053] According to a third aspect, a computer-readable storage medium is provided, on which a computer program is stored, and the computer program implements the method described in the first aspect when running.

[0054] The method and system for electromagnetic anti-interference at sea based on software radio broadband ad hoc network provided by the embodiment of the present invention can obtain a state detection report of a marine broadband ad hoc network to be optimized for electromagnetic anti-interference, and obtain an abnormal communication state identification record corresponding to the state detection report of the marine broadband ad hoc network; perform environmental factor variable mining on the state detection report of the marine broadband ad hoc network to obtain event environmental factor variables of the communication state reporting event in the state detection report of the marine broadband ad hoc network; perform electromagnetic interference factor mining on the abnormal communication state identification record to obtain electromagnetic interference factor variables of the abnormal mask text block in the abnormal communication state identification record; perform feature aggregation on the event environmental factor variables and the electromagnetic interference factor variables to obtain global marine communication factor variables corresponding to each aggregation channel of the communication state reporting event; perform feature linkage operation on the global marine communication factor variables of different aggregation channels to obtain marine communication linkage factor variables corresponding to the state detection report of the marine broadband ad hoc network; perform electromagnetic anti-interference decision analysis on the state detection report of the marine broadband ad hoc network based on the marine communication linkage factor variables to generate an electromagnetic anti-interference decision plan corresponding to the state detection report of the marine broadband ad hoc network. In view of the fact that the present invention can perform feature aggregation on the event environment factor variables of the maritime broadband self-organizing network status detection report and the electromagnetic interference factor variables corresponding to the abnormal communication status identification record, and obtain the global maritime communication factor variables corresponding to each aggregation channel of the communication status reporting event in the maritime broadband self-organizing network status detection report, it is possible to perform feature linkage operations on the global maritime communication factor variables of different aggregation channels, and obtain the maritime communication linkage factor variables with more interactive features in the maritime broadband self-organizing network status detection report, so that the electromagnetic anti-interference decision can be made for the maritime broadband self-organizing network status detection report based on the maritime communication linkage factor variables, thereby improving the generation accuracy and pertinence of the electromagnetic anti-interference decision plan corresponding to the maritime broadband self-organizing network status detection report. BRIEF DESCRIPTION OF THE DRAWINGS

[0055] In order to more clearly illustrate the technical solutions of the embodiments of the present invention, the drawings required for use in the embodiments are briefly introduced below. It should be understood that the following drawings only show certain embodiments of the present invention and therefore should not be regarded as limiting the scope. For ordinary technicians in this field, other related drawings can be obtained based on these drawings without creative work.

[0056] Figure 1 The present invention provides a flowchart of a method for processing electromagnetic interference at sea based on software radio technology. DETAILED DESCRIPTION

[0057] In order to better understand the above technical scheme, the technical scheme of the present invention is described in detail below through the accompanying drawings and specific embodiments. It should be understood that the embodiments of the present invention and the specific features in the embodiments are detailed descriptions of the technical scheme of the present invention, rather than limitations on the technical scheme of the present invention. In the absence of conflict, the embodiments of the present invention and the technical features in the embodiments can be combined with each other.

[0058] Figure 1 A method for marine electromagnetic anti-interference based on software radio broadband ad hoc network is shown, which is applied to a marine electromagnetic anti-interference processing system. The method includes the following steps 110 to 160.

[0059] Step 110: obtaining a state detection report of a marine broadband ad hoc network to be optimized for electromagnetic interference resistance, and obtaining an abnormal communication state identification record corresponding to the state detection report of the marine broadband ad hoc network.

[0060] Step 120: mining environmental factor variables on the maritime broadband ad hoc network status detection report to obtain event environmental factor variables of the communication status reporting event in the maritime broadband ad hoc network status detection report.

[0061] Step 130: Perform electromagnetic interference factor mining on the abnormal communication state identification record to obtain electromagnetic interference factor variables of the abnormal mask text block in the abnormal communication state identification record.

[0062] Step 140: feature-aggregate the event environment factor variables and the electromagnetic interference factor variables to obtain global maritime communication factor variables corresponding to each aggregated channel of the communication status reporting event.

[0063] Step 150: Perform feature linkage operation on the global maritime communication element variables of different aggregated channels to obtain maritime communication linkage element variables corresponding to the maritime broadband ad hoc network status detection report.

[0064] Step 160: Based on the maritime communication linkage factor variables, an electromagnetic anti-interference decision analysis is performed on the maritime broadband self-organizing network status detection report to generate an electromagnetic anti-interference decision plan corresponding to the maritime broadband self-organizing network status detection report.

[0065] The present invention can obtain a marine broadband self-organizing network status detection report to be optimized for electromagnetic anti-interference, and obtain an abnormal communication status identification record corresponding to the marine broadband self-organizing network status detection report; perform environmental factor variable mining on the marine broadband self-organizing network status detection report to obtain event environmental factor variables of a communication status reporting event in the marine broadband self-organizing network status detection report; perform electromagnetic interference factor mining on the abnormal communication status identification record to obtain electromagnetic interference factor variables of an abnormal mask text block in the abnormal communication status identification record; perform feature aggregation on the event environmental factor variables and the electromagnetic interference factor variables to obtain global marine communication factor variables corresponding to each aggregation channel of the communication status reporting event; perform feature linkage operation on the global marine communication factor variables of different aggregation channels to obtain marine communication linkage factor variables corresponding to the marine broadband self-organizing network status detection report; perform electromagnetic anti-interference decision analysis on the marine broadband self-organizing network status detection report based on the marine communication linkage factor variables to generate an electromagnetic anti-interference decision plan corresponding to the marine broadband self-organizing network status detection report. In view of the fact that the present invention can perform feature aggregation on the event environment factor variables of the maritime broadband self-organizing network status detection report and the electromagnetic interference factor variables corresponding to the abnormal communication status identification record, and obtain the global maritime communication factor variables corresponding to each aggregation channel of the communication status reporting event in the maritime broadband self-organizing network status detection report, it is possible to perform feature linkage operations on the global maritime communication factor variables of different aggregation channels, and obtain the maritime communication linkage factor variables with more interactive features in the maritime broadband self-organizing network status detection report, so that the electromagnetic anti-interference decision can be made for the maritime broadband self-organizing network status detection report based on the maritime communication linkage factor variables, thereby improving the generation accuracy and pertinence of the electromagnetic anti-interference decision plan corresponding to the maritime broadband self-organizing network status detection report.

[0066] The above steps are described in detail below.

[0067] In step 110, the maritime broadband ad hoc network status detection report is a detailed detection report that reflects the operating status of the maritime broadband ad hoc network. This may include but is not limited to key information such as network performance data, communication link quality, node status, etc. This report can provide network administrators with insights into network health and potential problems. The abnormal communication status identification record is a detailed record of any abnormal communication status found in the maritime broadband ad hoc network. For example, if a network node is suddenly lost, or the quality of a link drops abnormally, these situations will be recorded. These records can help identify and solve problems that may affect network performance.

[0068] Taking step 110 as an example, a maritime broadband self-organizing network operator collects a large amount of network performance data and generates a status detection report every day. At the same time, the marine electromagnetic anti-interference processing system will automatically monitor the network, and once any abnormal communication status is found, a corresponding identification record will be generated. For example, on a certain day, the operator found that the performance of an important communication link had degraded. First, the status detection report of the day will be checked, and it will be found that the performance of this link has been declining in the past few hours. Then, the abnormal communication status identification record will be checked, and it will be found that the system has automatically identified the abnormality of this link a few hours ago. In this way, this information can be used for in-depth analysis and research to determine the root cause of the problem and find the best solution. If the problem is caused by electromagnetic interference, this data can also be used to optimize the network and reduce the impact of electromagnetic interference.

[0069] In the maritime broadband ad hoc network environment, status detection reporting and abnormal communication status identification and recording are two crucial parts. The following is a detailed explanation of these two functions through specific examples.

[0070] The maritime broadband ad hoc network status detection report is mainly used to monitor and record the operating status of the network. For example, a ship can periodically send its current communication status to the base station by generating a status detection report. Such a report may include the following information:

[0071] The vessel's position and speed;

[0072] The frequency and power currently used;

[0073] Transmission rate and bit error rate;

[0074] The quality of the current network connection (e.g., signal strength, signal-to-noise ratio, etc.);

[0075] Such reports can help operators understand the operation of the network in real time and make adjustments as needed.

[0076] The abnormal communication status identification record is a tool used to monitor and record any abnormal conditions that occur in the network. For example, if a ship suddenly loses connection with the base station, the system will automatically generate an abnormal communication status identification record. This record may include the following information:

[0077] When and where the anomaly occurred;

[0078] The state of the vessel when the anomaly occurred (e.g., position, speed, frequency, power, etc.);

[0079] The network status when the anomaly occurred (for example, transmission rate, bit error rate, connection quality, etc.).

[0080] Such records can help operators quickly identify and solve problems to ensure stable network operation. At the same time, by analyzing these records, operators can also discover potential problems and further optimize the network.

[0081] In step 120, environmental factor variable mining is a data analysis technology that aims to extract important parameters or indicators related to the communication environment from a large number of maritime broadband ad hoc network status detection reports. These parameters or indicators may include weather conditions, sea conditions, ship positions, etc. A communication status reporting event is a situation that occurs during network operation, which may involve changes in network performance, link quality, node status, etc., and these changes are recorded and reported. Event environmental factor variables are important parameters or indicators related to the event environment mined from the above-mentioned communication status reporting events.

[0082] Taking step 120 as an example, after obtaining a status detection report, it can be found that the performance of a certain communication link has degraded. First, this matter will be regarded as a "communication status reporting event". Then, "environmental factor variable mining" will be carried out to try to find out the environmental factors that may affect the performance of the communication link, such as whether the weather is bad, whether the sea conditions are stable, whether the ship is moving, etc. These important parameters or indicators related to the event environment are the so-called "event environment factor variables". By mining and analyzing these variables, we can better understand the causes of communication status reporting events, and then formulate more effective electromagnetic interference prevention strategies.

[0083] Event environment variables or feature vectors are key elements that describe and define the environment in which an event occurs. They are usually expressed in numerical form, and these numerical values ​​can be calculated and analyzed to provide a deeper understanding of the event.

[0084] For example, in a maritime broadband ad hoc network, if a communication link quality degradation event is being processed, there may be the following environmental factor variables:

[0085] Signal strength (dBm) is a basic measure of wireless signal strength, usually measured in decibel milliwatts (dBm). For example, signal strength may be between -30dBm (very strong signal) and -90dBm (very weak signal);

[0086] Signal-to-noise ratio (dB): The signal-to-noise ratio is the ratio between the signal and the noise, usually expressed in decibels (dB). A higher signal-to-noise ratio indicates a better signal quality. For example, the signal-to-noise ratio may be between 20dB (excellent) and 0dB (poor);

[0087] Frequency (Hz) is the frequency used by the wireless signal and is measured in Hertz (Hz). For example, a ship might communicate on the 2.4GHz (which is 2400MHz or 2.4x10^9Hz) or 5GHz band.

[0088] The rate (Mbps) is the speed at which data is transmitted, usually measured in megabits per second (Mbps). For example, the rate may be between 1Mbps (slow) and 1000Mbps (i.e. 1Gbps, very fast);

[0089] These event environment factor variables can help better understand and analyze the problem and find possible solutions. For example, if the signal strength decreases, you may need to increase the transmission power; if the signal-to-noise ratio decreases, you may need to find a cleaner frequency band for communication; if the rate decreases, you may need to optimize the network settings or upgrade the equipment.

[0090] In data analysis and machine learning, feature vectors are a common representation method. For event environment factor variables, each factor can be regarded as a dimension, and then all factors are combined into a vector. For example, there are the following environmental factors:

[0091] Weather conditions (such as temperature, humidity, wind speed, etc.);

[0092] Sea conditions (such as wave height, current direction, etc.);

[0093] Vessel status (such as position, speed, heading, etc.);

[0094] Communication link status (such as signal strength, bit error rate, etc.).

[0095] You can assign specific values ​​to these elements and combine them into a vector. For example:

[0096] Weather conditions: temperature 25°C, humidity 80%, wind speed 10m / s;

[0097] Sea conditions: wave height 2m, current direction north;

[0098] Ship status: position (120.75, 30.5), speed 20 knots, heading east;

[0099] Communication link status: signal strength -70dBm, bit error rate 0.01.

[0100] Then, the characteristic vector of this event environment factor variable may be expressed as: [25, 80, 10, 2, "North", 120.75, 30.5, 20, "East", -70, 0.01].

[0101] In this way, a feature vector can be used to represent an event environment factor variable, which is convenient for further analysis and processing. Note that each element (or dimension) in this vector needs to have a clear definition and value range to accurately reflect the state of the environment.

[0102] In step 130, electromagnetic interference factor mining is a data analysis technology that is mainly used to extract electromagnetic interference related factors that may cause communication anomalies from abnormal communication status identification records. These factors may include electromagnetic frequency, interference source type, interference intensity, etc. The abnormal mask text block is a description or mark of a specific abnormal event in the abnormal communication status identification record, which can help better understand what kind of abnormality has occurred and find possible electromagnetic interference factors. Electromagnetic interference factor variables are important parameters or indicators related to electromagnetic interference mined from the above-mentioned abnormal mask text block.

[0103] Taking step 130 as an example, there is an abnormal communication state identification record, in which a communication link is suddenly interrupted. In this abnormal mask text block, some descriptive information may be included, such as "the communication link is suddenly interrupted on the 1900MHz frequency band, and the signal strength drops from -70dBm to -90dBm for 5 minutes".

[0104] Through electromagnetic interference factor mining, some key electromagnetic interference factor variables can be found, such as:

[0105] Interference frequency: 1900MHz;

[0106] Signal strength drop: 20dBm;

[0107] Interference duration: 5 minutes.

[0108] These electromagnetic interference factor variables can help better understand the causes of communication link interruptions and develop corresponding anti-interference strategies.

[0109] In other examples, the electromagnetic interference factor variable is used to describe various factors that may affect the quality of communication. These factors may include:

[0110] The type of interference source (e.g., antenna, power supply, radar, etc.);

[0111] The strength of the interference source (such as power, voltage, etc.);

[0112] The distance between the interference source and the receiver;

[0113] Interference frequency;

[0114] Duration of the disturbance.

[0115] Likewise, specific numerical values ​​can be assigned to these elements and combined into a vector.

[0116] For example:

[0117] Interference source: antenna;

[0118] Interference intensity: power 100W, voltage 220V;

[0119] Distance: 500m;

[0120] Interference frequency: 2.4GHz;

[0121] Duration: 10 seconds.

[0122] Then, the eigenvector of this electromagnetic interference factor variable may be expressed as: ["antenna", 100, 220, 500, 2.4, 10].

[0123] This feature vector provides a comprehensive description of the electromagnetic interference, enabling a better understanding of the nature of the interference source and appropriate measures to be taken to reduce its impact on communications.

[0124] In step 140, feature aggregation is a data processing technique that combines multiple related features or parameters to form a larger feature set or feature vector. This helps to understand and analyze problems from different perspectives. Aggregation channels are a link in the feature aggregation process, and each channel corresponds to a set of aggregated features. For example, all features related to signal strength can be placed in one channel, and all features related to frequency can be placed in another channel. The global maritime communication element variable is a set of feature variables obtained after feature aggregation, which reflects the global state of maritime communications, including environmental factors and electromagnetic interference factors.

[0125] Taking step 140 as an example, a set of event environment factor variables and electromagnetic interference factor variables have been obtained from steps 120 and 130. These variables may include weather conditions, sea conditions, ship status, interference frequency, signal strength drop, interference duration, etc.

[0126] Then, feature aggregation can be performed. For example, all variables related to environmental conditions (weather, sea conditions, etc.) can be put into one aggregation channel, and all variables related to electromagnetic interference (interference frequency, signal strength drop, etc.) can be put into another aggregation channel.

[0127] In this way, the global maritime communication element variables corresponding to each aggregated channel are obtained. These variables provide more comprehensive information, which is helpful for understanding and analyzing the status of maritime communications, so as to formulate more effective electromagnetic anti-interference strategies.

[0128] The global maritime communication element variable is a comprehensive description of the entire maritime broadband ad hoc network status, which can include environmental factors, communication equipment status, electromagnetic interference factors and other information.

[0129] For example:

[0130] Environmental factors: weather conditions (such as temperature, humidity, wind speed, etc.), sea conditions (such as wave height, current direction, etc.);

[0131] Communication equipment status: equipment location, usage frequency, signal strength, bit error rate, etc.;

[0132] Electromagnetic interference factors: interference source type, interference intensity, interference frequency, interference duration, etc.

[0133] In order to construct the feature vector of the global maritime communication element variable, each of the above elements can be assigned a specific value and these values ​​can be combined into a vector. For example:

[0134] Environmental factors: temperature 25℃, humidity 80%, wind speed 10m / s, wave height 2m, current direction north;

[0135] Communication device status: location (120.75, 30.5), frequency 2.4GHz, signal strength -70dBm, bit error rate 0.01;

[0136] Electromagnetic interference factors: interference source radar, interference intensity 100W, interference frequency 1900MHz, interference duration 5 minutes.

[0137] Then, the feature vector of this global maritime communication element variable may be expressed as: [25, 80, 10, 2, "North", 120.75, 30.5, 2.4, -70, 0.01, "Radar", 100, 1900, 5].

[0138] This feature vector provides a comprehensive description of the entire maritime communication environment, facilitating more accurate analysis and decision making.

[0139] In step 150, the feature linkage operation refers to combining the feature sets of different aggregated channels through a certain algorithm or logic to reveal the interaction and correlation between different features. This operation aims to find the intrinsic connection between features, thereby providing a deeper understanding of the problem. The maritime communication linkage factor variables are a set of comprehensive feature variables obtained through the feature linkage operation, which can reflect how different factors synergistically affect the communication status of the maritime broadband ad hoc network.

[0140] For example, it is known that the global marine communication factor variables obtained from step 140 include feature vectors A related to environmental conditions (such as temperature, humidity, wind speed, etc.) and feature vectors B related to electromagnetic interference (such as interference frequency, signal strength drop, etc.). It is now necessary to perform feature linkage operations to determine how these different feature sets work together on the communication quality of the marine broadband ad hoc network.

[0141] The feature linkage operation may include the following steps:

[0142] Analyze the impact of each environmental parameter in the feature vector A on the communication quality;

[0143] Analyze the influence of electromagnetic interference parameters in characteristic vector B on communication quality;

[0144] Explore whether there is an interaction between environmental conditions and electromagnetic interference, and how this interaction affects the stability of the communication link;

[0145] Use statistical analysis methods (such as regression analysis, principal component analysis, etc.) or machine learning algorithms (such as decision trees, random forests, etc.) to discover potential correlations between features.

[0146] Through such linkage operations, it may be found that under high humidity conditions, the impact of electromagnetic interference will be more significant, resulting in an increased probability of communication interruption. Therefore, the final marine communication linkage factor variable may contain a new variable C that measures the interaction between humidity and electromagnetic interference affecting communication quality.

[0147] The result of this process may be a new feature vector C, which contains the information of the original feature set and adds new insights into the interaction between the two. For example, C may include temperature, humidity, wind speed, interference frequency, signal strength drop, and an additional "humidity-interference interaction term". Such linkage variables can provide more accurate data support for subsequent electromagnetic anti-interference decision analysis.

[0148] In other examples, the maritime communication linkage factor variables are obtained by performing feature linkage operations on the global maritime communication factor variables of different aggregated channels. This operation aims to reveal the relationship between different features and how they jointly affect the state of the maritime broadband ad hoc network.

[0149] For example, the eigenvectors of the global maritime communication element variables of the two aggregated channels have been obtained from the previous steps:

[0150] Environmental and communication equipment status feature vector: [25°C, 80% humidity, 10m / s wind speed, 2m wave height, 120.75 longitude, 30.5 latitude, 2.4GHz frequency, -70dBm signal strength];

[0151] Electromagnetic interference factor feature vector: ["radar", 100W interference intensity, 1900MHz interference frequency, 5 minutes interference duration].

[0152] In a feature linkage operation, one might look for correlations or patterns between these features, such as whether the presence of radar is always accompanied by a drop in signal strength at a particular frequency, or whether communication links are more susceptible to interference under certain weather conditions.

[0153] Through these analyses, some new linkage features may be obtained, such as:

[0154] Correlation between radar jamming and signal strength degradation;

[0155] The impact of weather conditions on interference intensity;

[0156] Relationship between wave height and communication link quality.

[0157] For example, the following linkage relationships were found through analysis:

[0158] When radar interference occurs, if the wind speed exceeds 8m / s at the same time, the probability of signal strength decreasing by more than 20dBm increases;

[0159] When the humidity is below 60%, the signal quality of the 2.4GHz frequency is significantly better than when the humidity is above 80%;

[0160] If the wave height exceeds 3m, the bit error rate increases by 0.02 on average.

[0161] Based on these linkage relationships, a new feature vector can be constructed, which is expressed as the maritime communication linkage factor variable: [signal strength drop-radar interference correlation, weather conditions-interference intensity impact, wave height-communication quality relationship];

[0162] The specific numerical representation may be as follows: [1 (1 when wind speed > 8m / s, otherwise 0), 1 (1 when humidity < 60%, otherwise 0), 0.02 (increment when current wave height > 3m)].

[0163] Therefore, the new feature vector might be: [1, 1, 0.02].

[0164] This eigenvector is a simplified representation of the variables of the maritime communication linkage elements, which can be used to further analyze, predict or optimize the performance of the maritime broadband ad hoc network. The actual situation will be more complicated and require more data and advanced analysis techniques to accurately define these linkage element variables.

[0165] In step 160, electromagnetic anti-interference decision analysis is a data analysis-based method for evaluating how to optimize the electromagnetic environment of the maritime broadband self-organizing network to improve communication quality and stability. The analysis is based on the maritime communication linkage factor variables mined and linked in the previous steps. The electromagnetic anti-interference decision plan is the output result of the electromagnetic anti-interference decision analysis, which includes a series of recommended strategies or measures to reduce or eliminate the impact of electromagnetic interference on the communication quality of the maritime broadband self-organizing network.

[0166] For example, a series of maritime communication linkage factor variables are obtained through step 150, which indicate that under certain environmental conditions (such as high humidity weather), electromagnetic interference in a specific frequency band (such as interference generated by radio equipment of nearby ships) will significantly increase the risk of communication failure.

[0167] Next, when conducting an electromagnetic interference decision analysis, the following steps may be taken:

[0168] Use statistical models or machine learning algorithms to evaluate the impact of different interference factors on the stability of communication links based on the variables of maritime communication linkage elements;

[0169] Identify the most influential interference sources and analyze potential solutions to mitigate these interferences, such as changing the operating frequency of radio equipment and enhancing signal coding technology;

[0170] Based on the analysis results, a variety of possible solutions are formulated, each of which includes specific operational steps and expected results.

[0171] Finally, the EMI decision plan may include the following recommendations:

[0172] When high humidity weather is forecast, adjust the frequency of communication equipment in the ad hoc network in advance to avoid known interference frequency bands;

[0173] Increase signal redundancy in critical communication links and use more robust error correction coding to resist data loss caused by interference;

[0174] For identified major interference sources, coordinate communication frequency management with other ships to reduce frequency conflicts;

[0175] Implement a dynamic frequency selection mechanism to enable the network to adjust the communication channel in real time to adapt to changes in the electromagnetic environment.

[0176] Through the above plan, the maritime broadband self-organizing network will be able to more effectively combat electromagnetic interference and maintain stable communication quality.

[0177] The beneficial effects of this technical solution can be summarized from the following aspects:

[0178] Systematic analysis and optimization: From obtaining status detection reports to generating final decision plans, the entire process covers the entire process of data collection, analysis and decision making. This systematic approach ensures that all relevant factors are taken into account when optimizing electromagnetic interference resistance;

[0179] Data-driven decision-making: By mining environmental factor variables and electromagnetic interference factor variables, as well as subsequent feature aggregation and linkage operations, it ensures that decisions are based on sufficient data analysis. This helps reduce intuition or experience-driven uncertainties and improves the accuracy and effectiveness of decision-making;

[0180] Deeper insights into communication problems: By conducting specialized analysis of abnormal communication status in the maritime broadband ad hoc network status detection report, we can gain a deeper understanding of the causes of communication problems and develop targeted solutions;

[0181] Enhanced electromagnetic anti-interference capability: The electromagnetic anti-interference decision analysis in step 160 relies on the detailed information obtained in the previous steps, making the final electromagnetic anti-interference plan more specific and executable, and enhancing the stability and reliability of the marine broadband ad hoc network in a complex electromagnetic environment;

[0182] Improve network performance: By continuously monitoring the network status and adjusting the optimization strategy in a timely manner, the overall performance of the network can be significantly improved, including communication quality, data transmission rate and network stability;

[0183] Preventive maintenance and fault response: The solution can help identify potential sources of electromagnetic interference and communication vulnerabilities, provide a basis for preventive maintenance, and quickly respond to problems when they occur;

[0184] Scalability and adaptability: Although the solution is designed for maritime broadband ad hoc networks, its analysis and decision-making framework is universal and can be applied to other types of wireless networks and even similar problems in other industries;

[0185] Promote knowledge accumulation: By recording and analyzing data on the impact of various environmental and electromagnetic interference factors on communications, a valuable knowledge base can be accumulated for future predictive analysis and network design optimization.

[0186] In summary, electromagnetic anti-interference optimization of the maritime broadband self-organizing network is carried out through a series of steps. First, network status detection reports and abnormal communication status records are collected. Next, environmental factors and electromagnetic interference factor variables are mined, and these variables are feature aggregated to form global maritime communication factor variables. Then, feature linkage operations are performed to reveal the mutual influence between different factors and generate maritime communication linkage factor variables. Finally, decision analysis is performed based on these linkage variables to generate electromagnetic anti-interference decision plans. The whole process improves the stability and communication quality of the maritime broadband self-organizing network in the face of complex electromagnetic environments, making the network more robust, reducing communication failures, and ensuring the safety and efficiency of maritime operations.

[0187] In the application scenario of submarine cable operation scene communication, the stability and reliability of communication are crucial. Since submarine cable laying, maintenance and overhaul work usually takes place in deep sea or in waters far from land, traditional communication methods may be interfered by seawater, seabed topography and other marine environmental factors, resulting in reduced or interrupted communication quality. In order to solve this problem, marine electromagnetic anti-interference processing systems are widely used in the communication methods of submarine cable operation scenes.

[0188] At the beginning of the submarine cable operation, the marine electromagnetic anti-interference processing system will first obtain environmental data of the submarine cable operation area, including seawater temperature, salinity, seabed topography and other information, which are crucial for the subsequent optimization of communication methods. At the same time, the system will monitor the quality of the communication link in real time to detect and deal with potential electromagnetic interference in a timely manner.

[0189] During the operation, the marine electromagnetic anti-interference processing system will continuously analyze the communication signals and identify and suppress various electromagnetic interferences through advanced algorithms. For example, the system can identify signal attenuation caused by changes in seabed topography, or electromagnetic noise generated by surrounding ships, radars and other equipment, and take corresponding compensation measures to ensure the stable transmission of communication signals.

[0190] In addition, the system can dynamically adjust communication parameters based on real-time ocean environment data and communication quality monitoring results. For example, when the seawater temperature or salinity changes significantly, the system will automatically adjust the transmission power, frequency and other parameters to adapt to the current ocean environment and ensure the stability and reliability of communication.

[0191] It is worth mentioning that the offshore electromagnetic anti-interference processing system also has powerful fault diagnosis and recovery functions. Once a communication link fails or is severely interfered with, the system will immediately activate the emergency plan and try to restore communication. This may include switching to a backup communication link, enabling enhanced signal processing technology and other measures to ensure the continuity and safety of submarine cable operations.

[0192] It can be seen that the marine electromagnetic anti-interference processing system plays a vital role in the communication mode in the submarine cable operation scene. It can not only monitor and suppress various electromagnetic interferences in real time, but also dynamically adjust the communication parameters according to the changes in the marine environment to ensure the stability and reliability of communication. This is of great significance to ensure the safety and efficiency of submarine cable operations.

[0193] In some optional embodiments, step 140 performs feature aggregation on the event environment factor variables and the electromagnetic interference factor variables to obtain global maritime communication factor variables corresponding to each aggregated channel of the communication status reporting event, including steps 141 and 142.

[0194] Step 141, perform involvement feature aggregation on the event environment involved factor variables in the event environment factor variables and the electromagnetic interference factor variables to obtain the involved factor variable relationship network between the abnormal communication status identification record and the maritime broadband ad hoc network status detection report.

[0195] Step 142: Determine the global maritime communication element variable corresponding to each aggregation channel of the communication status reporting event based on the involved element variable relationship network and the stage event environment element variable in the event environment element variable.

[0196] In the above embodiment, the event environment involved factor variables refer to environmental factors associated with a communication status reporting event. For example, weather conditions, sea conditions, equipment location, etc. may all affect the occurrence of communication events. Involved feature aggregation is a special feature aggregation method that not only considers individual features, but also focuses on the relationship between features. For example, an electromagnetic interference factor may only significantly affect the communication quality under specific environmental conditions. The involved factor variable relationship network is a model or framework obtained through involved feature aggregation. It reveals the relationship between different factor variables and can help understand how various factors jointly affect the communication status.

[0197] For example, event environment factor variables (such as weather conditions, sea conditions) and electromagnetic interference factor variables (such as interference source type, interference intensity) have been obtained. In step 141, the involved feature aggregation will be performed. For example, it may be found that a certain type of electromagnetic interference will significantly affect the communication quality only when the humidity is high. In this way, a relationship network of involved factor variables is obtained, which tells how environmental conditions and electromagnetic interference jointly affect the communication status. Then, in step 142, the stage event environment factor variables in the event environment factor variables are considered, that is, the environmental factors that may appear at different stages. For example, more severe electromagnetic interference may occur at sunset. Based on this information, the global maritime communication factor variables corresponding to each aggregated channel can be determined.

[0198] In this way, by considering the relationship between various factors more deeply, a more accurate global marine communication element variable is provided. This helps to better understand the communication problem and develop more effective solution strategies. In addition, by involving feature aggregation and involving element variable relationship network, it is also possible to discover influencing factors that may have been overlooked before, thereby further improving the electromagnetic anti-interference capability of the marine broadband ad hoc network.

[0199] In some preferred embodiments, step 142 determines the global maritime communication element variables corresponding to each aggregation channel of the communication status reporting event based on the involved element variable relationship network and the stage event environment element variables in the event environment element variables, including steps 1421-1423.

[0200] Step 1421: perform knowledge mapping on the involved factor variable relationship network to obtain an involved factor mapping knowledge relationship network corresponding to the involved factor variable relationship network.

[0201] Step 1422: Perform knowledge mapping on the stage event environment element variables in the event environment element variables to obtain the stage event environment element mapping knowledge corresponding to the stage event environment element variables.

[0202] Step 1423: Determine the global maritime communication element variables corresponding to each aggregation channel of the communication status reporting event based on the involved element mapping knowledge relationship network and the stage event environment element mapping knowledge.

[0203] In the above embodiment, knowledge mapping is a process of converting raw data or information into knowledge representation. Knowledge mapping can help better understand and interpret raw data. The involved element mapping knowledge relationship network is the knowledge expression form of the involved element variable relationship network obtained through knowledge mapping. The stage event environment element mapping knowledge is the knowledge expression form of the stage event environment element variable obtained through knowledge mapping.

[0204] First, in step 1421, the involved factor variable relationship network is converted into an involved factor mapping knowledge relationship network through a knowledge mapping method. For example, the original involved factor variable relationship network may include associated information such as "radar interference" and "signal strength reduction". Through knowledge mapping, this information can be converted into a higher-level knowledge expression, such as "radar interference usually causes signal strength reduction".

[0205] Then, in step 1422, the stage event environment factor variables are knowledge mapped to obtain stage event environment factor mapping knowledge. For example, the original stage event environment factor variables may include "humidity 80%", and through knowledge mapping, this information can be converted into "when the humidity is high (such as 80%)".

[0206] Finally, in step 1423, the global marine communication element variables are determined by combining the knowledge relationship network of the involved element mapping and the knowledge of the stage event environment element mapping. For example, the following global variables may be obtained: "When the humidity is high (such as 80%), radar interference usually causes the signal strength to decrease."

[0207] In this way, by mapping and fusing multi-source data and information, more explanatory and operational global maritime communication element variables are formed, which helps to deeply understand the reasons for changes in communication status and thus formulate more effective electromagnetic anti-interference strategies.

[0208] In some preferred embodiments, the knowledge mapping of the involved factor variable relationship network to obtain the involved factor mapping knowledge relationship network corresponding to the involved factor variable relationship network includes: obtaining text fine-grained information of the abnormal communication status identification record; based on the text fine-grained information, disassembling the involved factor variable relationship network to obtain the involved factor variable relationship subnet corresponding to the involved factor variable relationship network; performing knowledge mapping on the involved factor variable relationship subnet to obtain the involved factor mapping knowledge relationship subnet corresponding to the involved factor variable relationship subnet; based on the involved factor mapping knowledge relationship subnet, determining the involved factor mapping knowledge relationship network corresponding to the involved factor variable relationship network.

[0209] In the above embodiment, the text fine-grained information refers to the detailed information extracted from the abnormal communication state identification record. This information may include specific interference sources, impact levels, occurrence time, etc. The involved factor variable relationship subnet is obtained by disassembling the involved factor variable relationship network. It represents the relationship between more specific and detailed factors. The involved factor mapping knowledge relationship subnet is obtained by performing knowledge mapping on the involved factor variable relationship subnet. It converts actual factor relationships into knowledge relationships, which is helpful for understanding and memory.

[0210] For example, abnormal communication status identification records have been obtained, and fine-grained text information has been extracted from them, such as "At sunset, due to increased humidity, a certain type of electromagnetic interference caused the communication quality to deteriorate." Then, based on this information, the involved factor variable relationship network is disassembled to obtain a more specific involved factor variable relationship subnetwork, such as "sunset-increased humidity-electromagnetic interference-deteriorated communication quality."

[0211] Next, knowledge mapping is performed on this subnet to transform it into an involved element mapping knowledge relationship subnet. For example, "sunset" may be mapped to "light intensity decreases", "increased humidity" may be mapped to "increased moisture content in the air", and so on. Finally, based on these subnets, the involved element mapping knowledge relationship network corresponding to the entire involved element variable relationship network can be determined.

[0212] Through this technical solution, the complex element relationships can be carefully disassembled and mapped to form a knowledge network that is easier to understand and remember. This helps to understand the causes of communication problems more deeply, so as to develop more effective solutions. In addition, this method can also discover influencing factors that may have been overlooked before, thereby further improving the electromagnetic anti-interference ability of the maritime broadband self-organizing network.

[0213] In some other preferred embodiments, the global maritime communication element variables corresponding to each aggregation channel of the communication status reporting event are determined based on the involved element mapping knowledge relationship network and the stage event environment element mapping knowledge, including: obtaining event description keywords corresponding to the event environment element variables, and obtaining event description summary word vectors corresponding to the involved element variable relationship network; performing linkage knowledge mining on the involved element mapping knowledge relationship network, the stage event environment element mapping knowledge, the event description keywords and the event description summary word vectors to obtain the global maritime communication element variables corresponding to each aggregation channel of the communication status reporting event.

[0214] Among them, event description keywords refer to important or core words used to describe the variables of the environmental elements of the communication status reporting event. Event description summary word vectors refer to a set of word vectors that can reflect the main features and content of the event by processing and transforming the relationship network of the involved element variables. Linkage knowledge mining is a data mining method that not only considers the impact of a single feature, but also focuses on the linkage relationship between features.

[0215] First, the event description keywords corresponding to the event environment factor variable are obtained. For example, for the event environment factor variable of bad sea conditions, the corresponding event description keywords may be "high waves and strong winds".

[0216] Secondly, obtain the event description summary word vector corresponding to the involved factor variable relationship network. For example, convert the knowledge expression such as "radar interference usually causes signal strength to decrease" into word vector form.

[0217] Then, the knowledge relationship network of involved element mapping, stage event environment element mapping knowledge, event description keywords and event description summary word vectors are linked for knowledge mining. For example, by analyzing this information, it may be found that the problem of signal strength reduction is more serious when the sea conditions are bad (high waves, strong winds) and there is radar interference.

[0218] Finally, based on this result, the global maritime communication element variables corresponding to each aggregated channel of the communication status reporting event are determined. For example, "when the sea conditions are bad and accompanied by radar interference, the signal strength drops seriously."

[0219] In this way, multiple data and information can be deeply integrated and mined to generate global maritime communication element variables with high interpretability and accuracy. This will help to better understand and predict changes in communication status, thereby formulating more effective electromagnetic anti-interference strategies and improving the stability and communication quality of maritime broadband ad hoc networks.

[0220] Under some exemplary design ideas, step 120 describes mining environmental factor variables of the maritime broadband self-organizing network status detection report to obtain event environmental factor variables of the communication status reporting event in the maritime broadband self-organizing network status detection report, including steps 121-122.

[0221] Step 121: mining environmental factor variables on the maritime broadband ad hoc network status detection report to obtain event type factor knowledge variables, event distribution factor knowledge variables and event impact factor knowledge variables of the communication status reporting event in the maritime broadband ad hoc network status detection report.

[0222] Step 122: Generate event environment element variables of the communication status reporting event in the maritime broadband ad hoc network status detection report based on the event type element knowledge variable, the event distribution element knowledge variable and the event impact element knowledge variable.

[0223] Among them, the event type factor knowledge variable refers to the knowledge expression form of specific event types related to communication status reporting events obtained by mining the maritime broadband self-organizing network status detection report. The event distribution factor knowledge variable refers to the knowledge expression form of the distribution characteristics of communication status reporting events in space or time obtained by mining the maritime broadband self-organizing network status detection report. The event impact factor knowledge variable refers to the knowledge expression form of the possible impact and degree of the communication status reporting event obtained by mining the maritime broadband self-organizing network status detection report.

[0224] First, in step 121, the marine broadband ad hoc network status detection report is mined for environmental factor variables. For example, it may be found that there are multiple communication status reporting events about "radar interference" in the report. These events mainly occur at sunset and usually lead to a decrease in communication quality. Therefore, the event type factor knowledge variable "radar interference", the event distribution factor knowledge variable "sunset" and the event impact factor knowledge variable "communication quality decrease" are obtained.

[0225] Then, in step 122, event environment factor variables are generated based on these knowledge variables. For example, a composite variable containing all the above information, or multiple simple variables each containing part of the information may be generated.

[0226] It can be seen that by mining the environmental factor knowledge variables with practical significance from the maritime broadband ad hoc network status detection report, a richer and deeper perspective is provided for understanding and solving communication problems. In addition, by integrating these knowledge variables into event environmental factor variables, the global characteristics of communication problems can be better grasped, thereby formulating more effective electromagnetic anti-interference strategies.

[0227] In some alternative embodiments, step 122 generates event environment element variables of the communication status reporting event in the maritime broadband self-organizing network status detection report based on the event type element knowledge variable, the event distribution element knowledge variable and the event impact element knowledge variable, including steps 1221-1222.

[0228] Step 1221: Aggregate information features of the event type factor knowledge variable, the event distribution factor knowledge variable, and the event impact factor knowledge variable to obtain the original event environment factor variable of each communication status reporting event in the maritime broadband ad hoc network status detection report.

[0229] Step 1222: perform an element variable splicing operation on the original event environment element variables corresponding to the different communication status reporting events to generate the event environment element variables of the communication status reporting event in the maritime broadband ad hoc network status detection report.

[0230] Among them, information feature aggregation is a data processing process that generates a more comprehensive and integrated feature description by summarizing and integrating multiple information features. The element variable splicing operation is a data processing method that combines different element variables to form a composite variable containing multiple element information.

[0231] In step 1221, information features of factor knowledge variables such as event type (e.g., radar interference), event distribution (e.g., interference sources are mainly concentrated in the northern waters), and event impact (e.g., causing a 30% decrease in communication quality) may be aggregated to generate original event environment factor variables.

[0232] In step 1222, the original event environment factor variables corresponding to different communication status reporting events may be spliced ​​together. For example, the information of radar interference, northern waters, and 30% decrease in communication quality may be spliced ​​together to form a more comprehensive event environment factor variable: "Radar interference in the northern waters causes a 30% decrease in communication quality."

[0233] Through this method, we can deeply understand the communication status reporting events from multiple dimensions (such as event type, distribution and impact), and generate more detailed and comprehensive event environment factor variables. This helps to more accurately identify and handle communication problems, thereby improving the stability and communication quality of maritime broadband ad hoc networks. At the same time, this method can also provide rich input information for subsequent data analysis and knowledge mining, further improving the intelligence level of the system.

[0234] In some optional embodiments, step 150 performs feature linkage operation on the global maritime communication element variables of different aggregated channels to obtain the maritime communication linkage element variables corresponding to the maritime broadband ad hoc network status detection report, including steps 151 and 152.

[0235] Step 151: Obtain the associated maritime communication element variable corresponding to the global maritime communication element variable.

[0236] Step 152: Based on the associated maritime communication element variables, feature linkage operation is performed on the global maritime communication element variables of different aggregated channels to obtain maritime communication linkage element variables corresponding to the maritime broadband ad hoc network status detection report.

[0237] In step 151, the associated marine communication element variables corresponding to the global marine communication element variables are obtained. For example, for the global variable "signal strength reduction caused by radar interference", the variables associated with it may be found, such as "humidity increase at sunset".

[0238] Next, in step 152, based on these associated variables, feature linkage operations are performed on the global maritime communication element variables of different aggregated channels. For example, it may be found that the signal strength drop caused by radar interference is more serious when the humidity increases at sunset. Therefore, a new maritime communication linkage element variable is obtained: "Increased humidity at sunset will aggravate the signal strength drop caused by radar interference."

[0239] In this way, the interactive influence relationship between the variables of maritime communication elements can be captured and analyzed from a global perspective, and more accurate and detailed variables of maritime communication linkage elements can be generated. This will help to deeply understand the complexity of the state of maritime broadband self-organizing networks, enhance the effectiveness of electromagnetic anti-interference strategies, and thus improve the quality and stability of maritime communications.

[0240] In some other possible embodiments, the performing of feature linkage operation on the global maritime communication element variables of different aggregated channels to obtain the maritime communication linkage element variables corresponding to the maritime broadband self-organizing network status detection report includes: calling the semantic feature processing branch in the natural language processing network to perform feature linkage operation on the global maritime communication element variables of different aggregated channels to obtain the maritime communication linkage element variables corresponding to the maritime broadband self-organizing network status detection report. The performing of electromagnetic anti-interference decision analysis on the maritime broadband self-organizing network status detection report based on the maritime communication linkage element variables to generate the electromagnetic anti-interference decision plan corresponding to the maritime broadband self-organizing network status detection report includes: calling the semantic feature processing branch in the natural language processing network to perform electromagnetic anti-interference decision analysis on the maritime broadband self-organizing network status detection report based on the maritime communication linkage element variables to generate the electromagnetic anti-interference decision plan corresponding to the maritime broadband self-organizing network status detection report.

[0241] The semantic feature processing branch is a component of the natural language processing network, which is responsible for extracting and processing semantically related features. In natural language processing tasks, semantic features include the meaning of words, phrases or sentences, which can help understand the content and meaning of the text.

[0242] For example, when performing feature linkage operations on global maritime communication element variables, the semantic feature processing branch may extract key information from each element variable, such as "radar interference", "northern waters", "communication quality decreased by 30%", etc., and identify the correlation between these information, such as the existence of a causal relationship between "radar interference" and "communication quality decreased".

[0243] Similarly, when generating electromagnetic anti-interference decision plans, the semantic feature processing branch may derive corresponding strategies based on the content of the maritime communication linkage factor variables, such as enhancing the anti-interference capability of the northern waters during periods of heavy radar interference.

[0244] By using the semantic feature processing branch, the semantic information in the marine communication element variables can be understood and utilized more accurately, and the effect of feature linkage operation and electromagnetic anti-interference decision analysis can be improved. This will help to better understand the actual situation of marine communications and develop more effective electromagnetic anti-interference strategies, thereby improving the stability and communication quality of marine broadband ad hoc networks.

[0245] With regard to the above-mentioned embodiment, first, the semantic feature processing branch in the natural language processing network is called to perform feature linkage operations on the global maritime communication element variables (such as "radar interference", "northern waters", "communication quality decreased by 30%) obtained from each aggregation channel to obtain maritime communication linkage element variables. For example, it may be found that when radar interference is severe and located in the northern waters, the problem of decreased communication quality is particularly prominent. Then, the semantic feature processing branch in the natural language processing network is called again, and based on this linkage element variable, an electromagnetic anti-interference decision analysis is performed on the marine broadband self-organizing network status detection report to generate an electromagnetic anti-interference decision plan. For example, it may be decided to improve the anti-radar interference capability of the northern waters during a specific time period of each day to cope with possible communication problems.

[0246] Through this technical solution, we can better explore and utilize the global information of maritime communications and further optimize the electromagnetic anti-interference strategy. At the same time, since the decision plan is generated based on the real-time communication status and environmental conditions, it can deal with various electromagnetic interference problems more accurately and effectively, thereby significantly improving the stability and communication quality of maritime broadband self-organizing networks.

[0247] In some possible embodiments, the calling of the semantic feature processing branch in the natural language processing network, based on the maritime communication linkage element variable, performs electromagnetic anti-interference decision analysis on the maritime broadband self-organizing network status detection report to generate an electromagnetic anti-interference decision plan corresponding to the maritime broadband self-organizing network status detection report, including: calling the semantic feature processing branch in the natural language processing network, based on the maritime communication linkage element variable, identifying the target electromagnetic interference characterization vector for the maritime broadband self-organizing network status detection report; based on the target electromagnetic interference characterization vector, performs electromagnetic anti-interference simulation on the abnormal communication status identification record to generate an electromagnetic anti-interference decision plan corresponding to the maritime broadband self-organizing network status detection report.

[0248] The target electromagnetic interference characterization vector refers to a set of values ​​used to represent the expected or target electromagnetic interference state. It can help understand and locate the specific situation of electromagnetic interference. Electromagnetic anti-interference simulation refers to simulating the electromagnetic environment under abnormal communication conditions based on the target electromagnetic interference characterization vector to predict and evaluate the effects of different anti-interference strategies.

[0249] First, the semantic feature processing branch in the natural language processing network is called to identify the target electromagnetic interference representation vector based on the maritime communication linkage factor variables, such as "the signal strength decrease caused by radar interference in the northern waters." For example, key features such as the frequency and intensity of radar interference may be identified.

[0250] Then, based on this characterization vector, electromagnetic anti-interference simulation is performed. For example, it may be simulated how much the communication quality can be improved under different anti-interference strategies, such as adjusting the signal frequency, enhancing the signal strength, etc. According to the simulation results, the corresponding electromagnetic anti-interference decision plan can be generated.

[0251] In this way, by deeply understanding the semantic information in the variables of maritime communication linkage elements, the state of electromagnetic interference can be accurately located and described, thus providing strong support for electromagnetic anti-interference strategies. Through electromagnetic anti-interference simulation, the effects of various strategies can be evaluated in advance, and the optimal strategy can be selected for implementation, thereby effectively improving the communication quality of maritime broadband ad hoc networks and improving their stability and reliability.

[0252] In some examples, the event environment factor variable is an AI knowledge factor variable obtained by calling a natural language processing network to perform environment factor variable mining on the maritime broadband self-organizing network status detection report. Based on this, before performing environment factor variable mining on the maritime broadband self-organizing network status detection report to obtain the event environment factor variable of the communication status reporting event in the maritime broadband self-organizing network status detection report, the method further includes steps 210 to 240.

[0253] Step 210: Obtain a network debugging example, wherein the network debugging example includes a maritime broadband self-organizing network status detection report learning example and an abnormal communication status identification record learning example corresponding to the maritime broadband self-organizing network status detection report learning example.

[0254] Step 220, call the element variable processing branch of the natural language processing network to be debugged, and based on the maritime broadband self-organizing network status detection report learning example and the abnormal communication status identification record learning example, obtain the sample global maritime communication element variable corresponding to each aggregation channel of the communication status reporting event in the maritime broadband self-organizing network status detection report learning example.

[0255] Step 230: Call the semantic feature processing branch of the natural language processing network to be debugged, and based on the sample global maritime communication element variables, perform a discrimination operation on the abnormal communication status identification record learning example to obtain an electromagnetic interference characterization vector prediction result corresponding to the abnormal communication status identification record learning example.

[0256] Step 240: Based on the electromagnetic interference characterization vector in the abnormal communication state identification record learning example and the electromagnetic interference characterization vector prediction result, the natural language processing network to be debugged is debugged to obtain the natural language processing network.

[0257] Among them, AI knowledge factor variables refer to environmental factor variables mined from the maritime broadband self-organizing network status detection report through artificial intelligence technology, such as natural language processing network. The network debugging example is a sample used to train and debug the natural language processing network, including a maritime broadband self-organizing network status detection report learning example and a corresponding abnormal communication status identification record learning example. The sample global maritime communication factor variable is a global maritime communication factor variable obtained by the factor variable processing branch of the natural language processing network based on the network debugging example. The electromagnetic interference characterization vector prediction result is a prediction result obtained by performing a discriminant operation on the sample global maritime communication factor variable by the semantic feature processing branch of the natural language processing network. The electromagnetic interference characterization vector is a feature vector in the abnormal communication status identification record learning example, which is used to characterize the characteristics of electromagnetic interference.

[0258] First, in step 210, network debugging examples are obtained, including learning examples of maritime broadband self-organizing network status detection reports and corresponding learning examples of abnormal communication status identification records. Then, in step 220, the element variable processing branch of the natural language processing network to be debugged is called, and based on these examples, the sample global maritime communication element variables corresponding to each communication status reporting event are obtained. In step 230, the semantic feature processing branch is called, and based on the sample global maritime communication element variables, the abnormal communication status identification record learning examples are discriminated, and the electromagnetic interference characterization vector prediction results are obtained. Finally, in step 240, the natural language processing network to be debugged is debugged according to the actual electromagnetic interference characterization vector and the prediction results to obtain the final natural language processing network.

[0259] In this way, through the use of AI knowledge factor variables and natural language processing networks, the information mined from the maritime broadband self-organizing network status detection report is more accurate and comprehensive, and can provide deeper insights. At the same time, through the use of network debugging examples, the performance of the natural language processing network can be continuously optimized and improved, so that it can better handle and analyze maritime communication problems in the future. This will help to formulate electromagnetic anti-interference strategies more effectively, thereby improving the stability and communication quality of maritime broadband self-organizing networks.

[0260] In some possible embodiments, before calling the semantic feature processing branch of the natural language processing network to be debugged as described in step 230, performing a discrimination operation on the abnormal communication status identification record learning example based on the sample global maritime communication element variable to obtain the electromagnetic interference characterization vector prediction result corresponding to the abnormal communication status identification record learning example, the method also includes steps 310-320.

[0261] Step 310: Obtain an associated state detection report learning example corresponding to the maritime broadband ad hoc network state detection report learning example.

[0262] Step 320: call the element variable processing branch of the natural language processing network to be debugged, and based on the associated status detection report learning example and the abnormal communication status identification record learning example, obtain the global maritime communication element variable associated learning example corresponding to each aggregated channel of the communication status reporting event in the maritime broadband self-organizing network status detection report learning example.

[0263] Based on this, the step 230 described in calling the semantic feature processing branch of the natural language processing network to be debugged, based on the sample global maritime communication element variable, performing a discrimination operation on the abnormal communication state identification record learning example, and obtaining the electromagnetic interference characterization vector prediction result corresponding to the abnormal communication state identification record learning example, includes: calling the semantic feature processing branch of the natural language processing network to be debugged, based on the sample global maritime communication element variable and the global maritime communication element variable associated learning example, performing a discrimination operation on the abnormal communication state identification record learning example, and obtaining the electromagnetic interference characterization vector prediction result corresponding to the abnormal communication state identification record learning example.

[0264] Among them, the associated state detection report learning example refers to other state detection report samples related to the given maritime broadband self-organizing network state detection report learning example. These associated reports can provide more contextual information to help understand and predict the communication status of a given report. The element variable processing branch refers to the part of the natural language processing network to be debugged that is specifically used to process and analyze element variables. The global maritime communication element variable associated learning example is an associated learning example obtained through the element variable processing branch based on the associated state detection report learning example and the abnormal communication state identification record learning example.

[0265] For example, there may already be a learning example of a maritime broadband ad hoc network status detection report, which reports the communication status at a certain moment. Then, other reports that are close in time to this report or have other similarities with it can be found, and these are the associated status detection report learning examples.

[0266] Next, the element variable processing branch in the natural language processing network is called to extract global maritime communication element variable association learning examples based on these associated reports and abnormal communication status recognition record learning examples. For example, it may be found that the degradation of communication quality in the northern waters is a common problem when radar interference is severe.

[0267] Finally, when predicting the electromagnetic interference characterization vector, not only the sample global maritime communication element variables of the given report are considered, but also this association learning example. This will make the prediction results more accurate and comprehensive.

[0268] In this way, by using the semantic feature processing branch and the element variable processing branch in the natural language processing network, combined with the association state detection report learning example and the abnormal communication state identification record learning example, a global maritime communication element variable association learning example is generated. This enables a better understanding and prediction of the state of maritime communications, thereby providing stronger support for decision-making. Specifically, this method can help more accurately predict the situation of electromagnetic interference, thereby formulating more effective electromagnetic anti-interference strategies and improving the communication quality and stability of maritime broadband ad hoc networks.

[0269] In summary, the electromagnetic interference status in the maritime communication scenario is identified and analyzed through the machine learning model, so that a targeted electromagnetic anti-interference processing solution is generated according to the identification and analysis results of the electromagnetic interference status, thereby improving the stability of maritime radio communications.

[0270] On the basis of the above, a marine electromagnetic anti-interference processing system is provided, comprising a processor and a memory communicating with each other, wherein the processor is used to retrieve a computer program from the memory and implement the above method by running the computer program.

[0271] Based on the above, a computer-readable storage medium is provided, on which a computer program is stored, and the computer program implements the above method when running.

[0272] The above are only embodiments of the present invention and are not intended to limit the present invention. For those skilled in the art, the present invention may have various modifications and variations. Any modification, equivalent substitution, improvement, etc. made within the spirit and principle of the present invention should be included in the scope of the claims of the present invention.

Claims

1. A method for marine electromagnetic interference resistance based on software radio broadband ad hoc network, characterized in that: Applied to a marine electromagnetic anti-interference processing system, the method comprises: Obtaining a state detection report of a maritime broadband self-organizing network to be optimized for electromagnetic interference resistance, and obtaining an abnormal communication state identification record corresponding to the state detection report of the maritime broadband self-organizing network; Performing environmental factor variable mining on the maritime broadband self-organizing network status detection report to obtain event environmental factor variables of the communication status reporting event in the maritime broadband self-organizing network status detection report; Performing electromagnetic interference factor mining on the abnormal communication state identification record to obtain electromagnetic interference factor variables of the abnormal mask text block in the abnormal communication state identification record; Performing feature aggregation on the event environment factor variables and the electromagnetic interference factor variables to obtain a global maritime communication factor variable corresponding to each aggregation channel of the communication status reporting event; Performing feature linkage operation on the global maritime communication element variables of different aggregation channels to obtain maritime communication linkage element variables corresponding to the maritime broadband ad hoc network status detection report; Based on the marine communication linkage element variables, an electromagnetic anti-interference decision analysis is performed on the marine broadband self-organizing network status detection report to generate an electromagnetic anti-interference decision plan corresponding to the marine broadband self-organizing network status detection report; The feature aggregation of the event environment factor variable and the electromagnetic interference factor variable to obtain the global maritime communication factor variable corresponding to each aggregation channel of the communication status reporting event includes: Performing involved feature aggregation on the event environment involved factor variables in the event environment factor variables and the electromagnetic interference factor variables to obtain an involved factor variable relationship network between the abnormal communication state identification record and the maritime broadband ad hoc network state detection report; Determine the global maritime communication element variable corresponding to each aggregation channel of the communication status reporting event based on the involved element variable relationship network and the stage event environment element variable in the event environment element variable; The determining of the global maritime communication element variable corresponding to each aggregation channel of the communication status reporting event based on the involved element variable relationship network and the stage event environment element variable in the event environment element variable includes: Performing knowledge mapping on the involved factor variable relationship network to obtain an involved factor mapping knowledge relationship network corresponding to the involved factor variable relationship network; Performing knowledge mapping on the stage event environment element variables in the event environment element variables to obtain the stage event environment element mapping knowledge corresponding to the stage event environment element variables; Determine the global maritime communication element variable corresponding to each aggregation channel of the communication status reporting event based on the involved element mapping knowledge relationship network and the stage event environment element mapping knowledge; The step of performing knowledge mapping on the involved factor variable relationship network to obtain an involved factor mapping knowledge relationship network corresponding to the involved factor variable relationship network includes: Acquiring text fine-grained information of the abnormal communication state identification record; Based on the text fine-grained information, the involved factor variable relationship network is disassembled to obtain the involved factor variable relationship subnetwork corresponding to the involved factor variable relationship network; Performing knowledge mapping on the involved factor variable relationship subnet to obtain an involved factor mapping knowledge relationship subnet corresponding to the involved factor variable relationship subnet; Based on the involved element mapping knowledge relationship subnetwork, determining the involved element mapping knowledge relationship network corresponding to the involved element variable relationship network; The determining of the global maritime communication element variable corresponding to each aggregation channel of the communication status reporting event based on the involved element mapping knowledge relationship network and the stage event environment element mapping knowledge includes: Obtaining event description keywords corresponding to the event environment element variables, and obtaining event description summary word vectors corresponding to the involved element variable relationship network; Performing linkage knowledge mining on the involved element mapping knowledge relationship network, the stage event environment element mapping knowledge, the event description keywords and the event description summary word vectors to obtain the global maritime communication element variables corresponding to each aggregation channel of the communication status reporting event; The environmental factor variables are mined for the maritime broadband self-organizing network status detection report to obtain event environmental factor variables of the communication status reporting event in the maritime broadband self-organizing network status detection report, including: Performing environmental factor variable mining on the maritime broadband ad hoc network status detection report to obtain event type factor knowledge variables, event distribution factor knowledge variables, and event impact factor knowledge variables of the communication status reporting event in the maritime broadband ad hoc network status detection report; Generate an event environment element variable of the communication status reporting event in the maritime broadband ad hoc network status detection report based on the event type element knowledge variable, the event distribution element knowledge variable and the event impact element knowledge variable; Wherein, the event environment element variable of the communication status reporting event in the maritime broadband self-organizing network status detection report is generated based on the event type element knowledge variable, the event distribution element knowledge variable and the event impact element knowledge variable, including: performing information feature aggregation on the event type element knowledge variable, the event distribution element knowledge variable and the event impact element knowledge variable to obtain the original event environment element variable of each communication status reporting event in the maritime broadband self-organizing network status detection report; performing element variable splicing operation on the original event environment element variables corresponding to different communication status reporting events to generate the event environment element variable of the communication status reporting event in the maritime broadband self-organizing network status detection report; The performing of feature linkage operation on the global maritime communication element variables of different aggregated channels to obtain the maritime communication linkage element variables corresponding to the maritime broadband ad hoc network status detection report includes: Obtaining an associated maritime communication element variable corresponding to the global maritime communication element variable; Based on the associated maritime communication element variables, a feature linkage operation is performed on the global maritime communication element variables of different aggregated channels to obtain a maritime communication linkage element variable corresponding to the maritime broadband ad hoc network status detection report; The performing a feature linkage operation on the global maritime communication element variables of different aggregated channels to obtain the maritime communication linkage element variables corresponding to the maritime broadband self-organizing network status detection report includes: calling a semantic feature processing branch in a natural language processing network to perform a feature linkage operation on the global maritime communication element variables of different aggregated channels to obtain the maritime communication linkage element variables corresponding to the maritime broadband self-organizing network status detection report; performing an electromagnetic anti-interference decision analysis on the maritime broadband self-organizing network status detection report based on the maritime communication linkage element variables to generate an electromagnetic anti-interference decision plan corresponding to the maritime broadband self-organizing network status detection report, including: calling a semantic feature processing branch in the natural language processing network to perform an electromagnetic anti-interference decision analysis on the maritime broadband self-organizing network status detection report based on the maritime communication linkage element variables to generate an electromagnetic anti-interference decision plan corresponding to the maritime broadband self-organizing network status detection report; Among them, the calling of the semantic feature processing branch in the natural language processing network, based on the maritime communication linkage element variable, performs electromagnetic anti-interference decision analysis on the maritime broadband self-organizing network status detection report to generate an electromagnetic anti-interference decision plan corresponding to the maritime broadband self-organizing network status detection report, including: calling the semantic feature processing branch in the natural language processing network, based on the maritime communication linkage element variable, identifying the target electromagnetic interference characterization vector for the maritime broadband self-organizing network status detection report; based on the target electromagnetic interference characterization vector, performs electromagnetic anti-interference simulation on the abnormal communication status identification record to generate an electromagnetic anti-interference decision plan corresponding to the maritime broadband self-organizing network status detection report; The event environment factor variable is an AI knowledge factor variable obtained by calling a natural language processing network to perform environment factor variable mining on the maritime broadband self-organizing network status detection report; before performing environment factor variable mining on the maritime broadband self-organizing network status detection report to obtain the event environment factor variable of the communication status reporting event in the maritime broadband self-organizing network status detection report, the method further includes: Obtaining a network debugging example, wherein the network debugging example includes a maritime broadband self-organizing network state detection report learning example and an abnormal communication state identification record learning example corresponding to the maritime broadband self-organizing network state detection report learning example; Calling the element variable processing branch of the natural language processing network to be debugged, based on the maritime broadband self-organizing network state detection report learning example and the abnormal communication state identification record learning example, obtaining the sample global maritime communication element variable corresponding to each aggregation channel of the communication state reporting event in the maritime broadband self-organizing network state detection report learning example; Calling the semantic feature processing branch of the natural language processing network to be debugged, based on the sample global maritime communication element variable, performing a discrimination operation on the abnormal communication state identification record learning example, and obtaining an electromagnetic interference characterization vector prediction result corresponding to the abnormal communication state identification record learning example; Based on the electromagnetic interference characterization vector in the abnormal communication state identification record learning example and the electromagnetic interference characterization vector prediction result, the natural language processing network to be debugged is debugged to obtain the natural language processing network; Among them, before calling the semantic feature processing branch of the natural language processing network to be debugged, performing a discrimination operation on the abnormal communication state identification record learning example based on the sample global maritime communication element variable, and obtaining the electromagnetic interference characterization vector prediction result corresponding to the abnormal communication state identification record learning example, the method also includes: obtaining the associated state detection report learning example corresponding to the maritime broadband self-organizing network state detection report learning example; calling the element variable processing branch of the natural language processing network to be debugged, based on the associated state detection report learning example and the abnormal communication state identification record learning example, obtaining the global maritime communication element variable associated learning example corresponding to each aggregated channel of the communication state reporting event in the maritime broadband self-organizing network state detection report learning example; Then, the calling of the semantic feature processing branch of the natural language processing network to be debugged performs a discrimination operation on the abnormal communication state identification record learning example based on the sample global maritime communication element variable to obtain a prediction result of the electromagnetic interference characterization vector corresponding to the abnormal communication state identification record learning example, including: calling the semantic feature processing branch of the natural language processing network to be debugged, performs a discrimination operation on the abnormal communication state identification record learning example based on the sample global maritime communication element variable and the global maritime communication element variable associated learning example to obtain a prediction result of the electromagnetic interference characterization vector corresponding to the abnormal communication state identification record learning example.

2. A marine electromagnetic anti-interference processing system, characterized in that: The marine electromagnetic anti-interference processing system includes a processor and a memory that communicate with each other, and the processor is used to call a computer program from the memory and implement the method described in claim 1 by running the computer program.

Images