A system for 5G local area network and satellite communication within a fleet

Abstract

The present invention relates to the field of radio communication technologies, and discloses a system for 5G local area network and satellite communication within a fleet, including: a data acquisition module, configured to acquire communication characteristic data, geographical locations, and control parameters of M ships in the fleet, and represent them by first characteristic data; a preprocessing and feature extraction module, configured to preprocess the first characteristic data to obtain second characteristic data, extract coupling features according to the second characteristic data, and form third characteristic data by combining the normalized coupling features and the second characteristic data; a communication quality score generation module, configured to generate communication quality scores of the M ships according to the third characteristic data of the M ships using a first scoring formula; a control parameter optimization module, configured to optimize the control parameters of the M ships according to an improved starfish optimization algorithm to obtain an optimal set of control parameters to maximize the communication performance of the fleet. The present invention realizes the maximization of the overall communication performance of the fleet.

Description

Technical Field

[0001] The present invention belongs to the technical field of radio communication, and particularly relates to a system for 5G local area network and satellite communication within a fleet. Background Art

[0002] With the continuous development of the shipping industry, efficient radio communication between fleets has become a key technology to improve the efficiency and safety of maritime operations. Currently, the bandwidth allocation ratio, transmission power, and channel coding rate that affect the quality of fleet radio communication usually rely on manual experience for adjustment. Although it can improve the communication quality to a certain extent, it cannot be adjusted in a timely manner according to the actual situation, lacking flexibility and accuracy, and cannot respond to changes in communication requirements in a timely manner, resulting in the communication quality within the fleet not reaching the optimal level. For example, the coverage area of the 5G network is limited, and the signal attenuation is serious when far from land, while satellite communication can maintain signal coverage in the open sea area, but is limited by bandwidth and latency. Therefore, a system that relies on static and empirical adjustments will lead to unstable communication quality and sub - optimal communication quality in practical applications. Summary of the Invention

[0003] The present invention provides a system for 5G local area network and satellite communication within a fleet, which solves the technical problem of relying on manual experience to adjust communication parameters in related technologies, lacking flexibility and real - time performance.

[0004] The present invention provides a system for 5G local area network and satellite communication within a fleet, including:

[0005] A data acquisition module, configured to collect communication characteristic data, geographical locations, and control parameters of M ships in the fleet, and represent them by first characteristic data. Among them, the m - th data unit of the first characteristic data represents the communication characteristic data, geographical location, and control parameters of the m - th ship. The communication characteristic data includes: signal strength, latency rate, throughput, bit error rate, network load, and packet loss rate. The control parameters include: bandwidth allocation ratio, transmission power, and channel coding rate. M and m are self - defined parameters;

[0006] A pre - processing and feature extraction module, configured to pre - process the first characteristic data to obtain second characteristic data, extract coupling features according to the second characteristic data, and form third characteristic data by combining the normalized coupling features and the second characteristic data;

[0007] A communication quality score generation module, configured to generate communication quality scores of M ships according to the third characteristic data of M ships using a first scoring formula;

[0008] A control parameter optimization module, configured to optimize the control parameters of M ships according to an improved starfish optimization algorithm to obtain an optimal set of control parameters to maximize the communication performance of the fleet.

[0009] Furthermore, preprocess the first feature data to obtain the second feature data. The specific steps include:

[0010] S201. For the missing values of the signal strength in the first feature data of the vessel, fill them with the signal strength of the vessel with the shortest Euclidean distance to this vessel;

[0011] S202. For the missing values of the features other than the signal strength in the first feature data of the vessel, fill them with the average value of the feature values other than the missing values;

[0012] S203. Use the maximum-minimum normalization method to normalize the processed first feature data to obtain the second feature data.

[0013] Furthermore, extract the coupling features according to the second feature data. The specific steps include:

[0014] S301. Extract the dynamic communication efficiency feature according to the second feature data. The calculation formula of the dynamic communication efficiency feature is: , where represents the dynamic communication efficiency feature, T represents the throughput, represents the packet loss rate, represents the delay rate, represents the bit error rate, represents the network load, represents the heading angle of the vessel, represents the bandwidth allocation ratio, represents the transmission power, represents the channel coding rate;

[0015] S302. Extract the channel stability feature according to the second feature data. The calculation formula of the channel stability feature is: , where represents the channel stability feature, S represents the signal strength, represents the weight coefficient of the packet loss rate on the channel stability, represents the cosine value of the vessel heading, represents the weight coefficient of the delay rate on the channel stability, represents the weight coefficient of the network load on the channel stability;

[0016] S303. Extract the network pressure index feature according to the second feature data. The calculation formula of the network pressure index feature is: , where represents the network pressure index feature, represents the correction coefficient of the bit error rate, represents the weight coefficient of the transmission power on the network pressure in the high-throughput scenario;

[0017] S304. Extract the channel utilization rate feature according to the second feature data. The calculation formula of the channel utilization rate feature is: , where represents the channel utilization rate feature, represents the logarithmic function weight coefficient of represents the first weight coefficient, represents the second weight coefficient.

[0018] Furthermore, generate the communication quality scores of M ships according to the third feature data of M ships using the first scoring formula. The first scoring formula is:

[0019] ;

[0020] where represents the communication quality score of the ship, N represents the number of features in the third feature data, a represents the feature index in the third feature data, represents the weight of the a-th feature, represents the feature value of the a-th feature in the third feature data.

[0021] Furthermore, calculate the weight of each feature in the third feature data in the first scoring formula according to the XGBoost algorithm. Specifically, use the split gain method to calculate the weight. The specific steps include:

[0022] S401. Configure the model parameters. The model parameters include: the depth of the tree, the learning rate, and the number of training rounds;

[0023] S402. Use the XGBoost algorithm to construct a decision tree. Specifically, starting from the root node, use the third feature data of M ships as the initial data, select the split feature according to the split gain formula, divide the initial data into left and right child nodes, and recursively construct subtrees until the depth of the tree is reached;

[0024] S403. For each tree in the XGBoost algorithm, find the node set that uses feature i as the split node, and calculate the split gain of each node to obtain the total split gain of feature i;

[0025] S404. Normalize the split gain of each feature to obtain the weight of each feature.

[0026] Furthermore, optimize the control parameters of M ships according to the improved starfish optimization algorithm to obtain the optimal control parameters. The specific steps include:

[0027] S501. Generate P initial solutions that meet the constraint conditions based on the uniform distribution;

[0028] S502. Construct an objective function based on the communication quality scores of M ships;

[0029] S503. Calculate the fitness value of each solution according to the objective function, and take the solution with the highest fitness value as the current optimal solution;

[0030] S504. Simulate the exploration behavior of starfish, and update each solution according to the current optimal solution using the first update formula. The first update formula is: , where represents the j-th solution in the (k + 1)-th iteration, represents the j-th solution in the k-th iteration, k represents the current iteration number, j represents the index of the solution, and are random numbers in the range of 0 to 1, represents the current optimal solution, represents a random angle;

[0031] S505. Simulate the regeneration stage of starfish, and update each solution using the second update formula. The second update formula is: , where K represents the maximum number of iterations;

[0032] S506. When the current iteration number reaches the maximum number of iterations, take the solution with the maximum current fitness value as the optimal set of control parameters, otherwise return to S503.

[0033] Furthermore, the objective function for improving the starfish optimization algorithm is:

[0034] ;

[0035] where F represents the fitness value of the solution, b represents the ship index, represents the communication quality score of the b-th ship.

[0036] The beneficial effects of the present invention are as follows: By collecting the communication characteristic data, geographical locations, and control parameters of each ship in the fleet, and combining with the improved starfish optimization algorithm, the present invention realizes the dynamic improvement of the communication quality of the fleet; this method ensures stable communication performance of the fleet and maximizes the overall communication efficiency of the fleet;

[0037] The present invention adopts the improved starfish optimization algorithm to optimize the control parameters of each ship in the fleet by simulating the exploration behavior and regeneration stage of starfish; the optimization process not only ensures the efficiency of global search but also makes fine adjustments in the local area, effectively improving the communication quality. BRIEF DESCRIPTION OF THE DRAWINGS

[0038] Figure 1 is a schematic diagram of the modules of a 5G local area network and satellite communication system inside a fleet according to the present invention. Detailed implementation manners

[0039] The subject matter described herein will now be discussed with reference to exemplary implementation manners. It should be understood that discussing these implementation manners is only to enable those skilled in the art to better understand and thus implement the subject matter described herein. Without departing from the scope of protection of the content of this specification, changes can be made to the functions and arrangements of the elements discussed. Each example can omit, substitute, or add various processes or components as needed. Additionally, the features described for some examples can also be combined in other examples.

[0040] As Figure 1 shown, a system for 5G local area network and satellite communication within a fleet includes:

[0041] A data acquisition module 101, configured to acquire communication characteristic data, geographical locations, and control parameters of M ships in the fleet, and represent them by first characteristic data. Among them, the m-th data unit of the first characteristic data represents the communication characteristic data, geographical location, and control parameters of the m-th ship. The communication characteristic data includes: signal strength, delay rate, throughput, bit error rate, network load, and packet loss rate. The control parameters include: bandwidth allocation ratio, transmission power, and channel coding rate. M and m are custom parameters;

[0042] A preprocessing and feature extraction module 102, configured to preprocess the first characteristic data to obtain second characteristic data, extract coupling features according to the second characteristic data, and form third characteristic data by combining the normalized coupling features and the second characteristic data;

[0043] A communication quality score generation module 103, configured to generate communication quality scores of M ships according to the third characteristic data of M ships using a first scoring formula;

[0044] A control parameter optimization module 104, configured to optimize the control parameters of M ships according to an improved starfish optimization algorithm to obtain an optimal set of control parameters to maximize the communication performance of the fleet.

[0045] In an embodiment of the present invention, communication characteristic data, geographical locations, and control parameters are acquired through the communication devices and network monitoring tools of unmanned ships. The geographical location is represented by a two-dimensional coordinate of longitude and latitude. The bandwidth allocation ratio represents the bandwidth ratio allocated to each ship, is represented by real number coding, and the sum of the bandwidth allocation ratios of M ships is 1. The transmission power is represented by real number coding, and the channel coding rate is represented by real number coding. Among them, the unit of signal strength is dBm, the unit of delay rate is ms, the unit of throughput is Mbps, and the unit of transmission power is W.

[0046] In one embodiment of the present invention, the first feature data is preprocessed to obtain second feature data. The specific steps include:

[0047] S201. For the missing values of the signal strength in the first feature data of the vessel, use the signal strength of the vessel with the shortest Euclidean distance to fill the missing values. Specifically, calculate the Euclidean distance between vessels based on the geographical locations of the vessels to obtain the vessel with the shortest distance to the vessel.

[0048] S202. For the missing values of the features other than the signal strength in the first feature data of the vessel, use the average value of the feature values other than the missing values for filling.

[0049] S203. Use the maximum - minimum normalization method to normalize the processed first feature data to obtain the second feature data.

[0050] In one embodiment of the present invention, coupled features are extracted based on the second feature data. The specific steps include:

[0051] S301. Extract the dynamic communication efficiency feature based on the second feature data. The calculation formula for the dynamic communication efficiency feature is: , where represents the dynamic communication efficiency feature, T represents the throughput, represents the packet loss rate, represents the delay rate, represents the bit error rate, represents the network load, represents the course angle of the vessel, represents the bandwidth allocation ratio, represents the transmission power, represents the channel coding rate, represents the effective data volume transmitted, represents the communication loss;

[0052] S302. Extract the channel stability feature based on the second feature data. The calculation formula for the channel stability feature is: , where represents the channel stability feature, S represents the signal strength, represents the weight coefficient of the packet loss rate on the channel stability, represents the cosine value of the vessel course, represents the weight coefficient of the delay rate on the channel stability, represents the weight coefficient of the network load on the channel stability, represents the interaction between the transmission power and the bandwidth, used to express the additional influence of resource utilization;

[0053] S303. Extract the network pressure index feature according to the second feature data. The calculation formula of the network pressure index feature is: , where represents the network pressure index feature, represents the correction coefficient of the bit error rate, represents the weight coefficient of the transmission power on the network pressure in the high-throughput scenario, represents the non-linear correction of the bit error rate, is used to simulate the influence of the transmission power on the network pressure in the high-throughput scenario;

[0054] S304. Extract the channel utilization rate feature according to the second feature data. The calculation formula of the channel utilization rate feature is: , where represents the channel utilization rate feature, represents the logarithmic function of the weight coefficient, represents the first weight coefficient, is used to represent the efficiency of the channel resources, represents the second weight coefficient, represents the compensation effect of the channel coding rate on the packet loss rate.

[0055] In an embodiment of the present invention, the coupling feature and the second feature data are combined to form the third feature data. Specifically, the third feature data includes: signal strength, delay rate, throughput, bit error rate, network load, packet loss rate, bandwidth allocation ratio, transmission power, channel coding rate, dynamic communication efficiency feature, channel stability feature, network pressure index feature, and channel utilization rate feature.

[0056] In an embodiment of the present invention, according to the third feature data of M ships, the communication quality scores of M ships are generated using the first scoring formula. The first scoring formula is:

[0057] ;

[0058] where represents the communication quality score of the ship, N represents the number of features in the third feature data, a represents the feature index in the third feature data, represents the weight of the a-th feature, represents the feature value of the a-th feature in the third feature data.

[0059] In an embodiment of the present invention, according to the XGBoost algorithm, the weight of each feature in the third feature data in the first scoring formula is calculated. Specifically, the split gain method is used to calculate the weight. The specific steps include:

[0060] S401. Configure model parameters, where the model parameters include: the depth of the tree, the learning rate, and the number of training rounds. Preferably, the depth of the tree is set to 6, the learning rate is set to 0.1, and the number of training rounds is set to 100 times;

[0061] S402. Use the XGBoost algorithm to construct a decision tree. Specifically, starting from the root node, use the third feature data of M ships as the initial data, select the splitting feature according to the splitting gain formula, divide the initial data into left and right child nodes, and recursively construct subtrees until the depth of the tree is reached. Among them, the splitting gain formula is: , represents the splitting gain, and represent the sum of gradients of the initial data in the left and right child nodes respectively, and represent the sum of second-order derivatives in the left and right child nodes respectively. G represents the sum of gradients of the current node, and H represents the sum of second-order derivatives of the current node. represents the complexity coefficient, which is used to control the complexity of the splitting gain;

[0062] S403. For each tree in the XGBoost algorithm, find the node set that uses feature i as the splitting node, and calculate the splitting gain of each node to obtain the total splitting gain of feature i;

[0063] S404. Normalize the splitting gain of each feature to obtain the weight of each feature.

[0064] In an embodiment of the present invention, optimize the control parameters of M ships according to the improved starfish optimization algorithm to obtain the optimal control parameters. The specific steps include:

[0065] S501. Generate P initial solutions that meet the constraint conditions based on a uniform distribution. Specifically, each solution represents a combination of the control parameters of M ships. The constraint conditions include: the sum of the bandwidth allocation ratios of M ships is 1; the transmission power and channel coding rate have their maximum and minimum values according to physical condition limitations;

[0066] For the transmission power and channel coding, generate P solutions using a uniform distribution and ensure that they are within the value range. For the bandwidth allocation ratio, randomly generate P solutions under the condition of meeting the constraint conditions;

[0067] S502. Construct an objective function based on the communication quality scores of M ships;

[0068] S503. Calculate the fitness value of each solution according to the objective function, and take the solution with the highest fitness value as the current optimal solution;

[0069] S504. Simulate the exploration behavior of starfish and update each solution using the first update formula according to the current optimal solution. The first update formula is: , where represents the j-th solution in the (k + 1)-th iteration, represents the j-th solution in the k-th iteration, k represents the current iteration number, j represents the index of the solution, and are random numbers in the range of 0 to 1, represents the current optimal solution, represents a random angle, simulating the movement pattern of starfish;

[0070] S505. Simulate the regeneration stage of starfish and update each solution using the second update formula. The second update formula is: , where K represents the maximum number of iterations. Preferably, K is taken as 200;

[0071] S506. When the current iteration number reaches the maximum number of iterations, take the solution with the maximum current fitness value as the optimal set of control parameters; otherwise, return to S503.

[0072] In an embodiment of the present invention, the objective function for improving the starfish optimization algorithm is:

[0073] ;

[0074] where F represents the fitness value of the solution, b represents the ship index, represents the communication quality score of the b-th ship.

[0075] In an embodiment of the present invention, the improved starfish optimization algorithm is used to optimize the control parameters of M ships. During the optimization process, the update amplitude of the solution is controlled by exponential decay, simulating the exploration behavior of starfish in the initial stage and the regeneration behavior of starfish in the later stage, ensuring efficient global search and fine local optimization, and finding the optimal set of control parameters for the entire fleet.

[0076] The above describes the embodiments of this embodiment, but this embodiment is not limited to the above specific implementation manners. The above specific implementation manners are merely illustrative and not restrictive. Under the inspiration of this embodiment, those of ordinary skill in the art can also make many forms, all of which fall within the protection scope of this embodiment.

Claims

1. A system for 5G local area network and satellite communication within a fleet, characterized in that: include: A data collection module is used to collect communication characteristic data, geographic location and control parameters of M ships in the fleet when communicating through the 5G network and satellite, and represent them through first characteristic data, wherein the mth data unit of the first characteristic data represents the communication characteristic data, geographic location and control parameters of the mth ship, the communication characteristic data includes: signal strength, delay rate, throughput, bit error rate, network load and packet loss rate, the control parameters include: bandwidth allocation ratio, transmission power and channel coding rate, and M and m are custom parameters; A preprocessing and feature extraction module, used for preprocessing the first feature data to obtain the second feature data, extracting the coupling feature according to the second feature data, and combining the normalized coupling feature and the second feature data into the third feature data; The specific steps of extracting the coupling feature according to the second feature data include: S301, extracting a dynamic communication efficiency feature according to the second feature data, wherein the calculation formula of the dynamic communication efficiency feature is: ,in, represents the dynamic communication efficiency characteristics, T represents the throughput, Indicates the packet loss rate, represents the delay rate, represents the bit error rate, Indicates the network load, Indicates the heading angle of the ship. Indicates the bandwidth allocation ratio. represents the transmission power, represents the channel coding rate; S302: extracting a channel stability feature according to the second feature data. The calculation formula of the channel stability feature is: ,in, represents the channel stability characteristic, S represents the signal strength, represents the weight coefficient of packet loss rate on channel stability, represents the cosine value of the ship's heading, represents the weight coefficient of delay rate on channel stability, Indicates the weight coefficient of network load on channel stability; S303, extracting a network pressure index feature according to the second feature data, the calculation formula of the network pressure index feature is: ,in, Represents the network stress index characteristics, Represents the correction factor of the bit error rate, Indicates the weight coefficient of transmission power on network pressure in high throughput scenarios; S304: extracting a channel utilization feature according to the second feature data. The calculation formula of the channel utilization feature is: ,in, represents the channel utilization characteristics, Represents a logarithmic function The weight coefficient of represents the first weight coefficient, represents the second weight coefficient; A communication quality score generating module, configured to generate communication quality scores of the M ships using a first scoring formula according to the third feature data of the M ships; The control parameter optimization module is used to optimize the control parameters of M ships according to the improved starfish optimization algorithm to obtain the optimal control parameter set.

2. A fleet internal 5G local area network and satellite communication system according to claim 1, characterized in that: Preprocessing the first characteristic data to obtain the second characteristic data includes: S201, for missing values ​​of signal strength in the first characteristic data of a ship, fill in the missing values ​​using the signal strength of the ship with the shortest Euclidean distance to the ship; S202, for missing values ​​of features other than signal strength in the first feature data of the vessel, use the average value of the feature values ​​other than the missing value to perform filling processing; S203, using a maximum-minimum normalization method to normalize the processed first feature data to obtain second feature data.

3. The system for 5G local area network and satellite communication within a fleet according to claim 1, characterized in that: The communication quality scores of the M ships are generated using the first scoring formula according to the third characteristic data of the M ships. The first scoring formula is: ； in, represents the communication quality score of the ship, N represents the number of features in the third feature data, a represents the feature index in the third feature data, represents the weight of the a-th feature, Represents the feature value of the ath feature in the third feature data.

4. A fleet internal 5G local area network and satellite communication system according to claim 3, characterized in that: The weight of each feature in the third feature data in the first scoring formula is calculated according to the XGBoost algorithm. Specifically, the weight is calculated using the split gain method. The specific steps include: S401, configure model parameters, which include: tree depth, learning rate, and training rounds; S402, using the XGBoost algorithm to construct a decision tree. Specifically, starting from the root node, the third feature data of the M ships is used as the initial data, a split feature is selected according to the split gain formula, the initial data is divided into left and right child nodes, and a subtree is recursively constructed until the depth of the tree is reached; S403, for each tree in the XGBoost algorithm, find a node set that uses feature i as a split node, and calculate the split gain of each node to obtain a total split gain of feature i; S404, normalize the split gain of each feature to obtain the weight of each feature.

5. The system for 5G local area network and satellite communication within a fleet according to claim 1, characterized in that: The control parameters of M ships are optimized according to the improved starfish optimization algorithm to obtain the optimal control parameters. The specific steps include: S501, generating P initial solutions that meet the constraint conditions based on uniform distribution; S502, constructing an objective function according to the communication quality scores of the M ships; S503, calculating the fitness value of each solution according to the objective function, and taking the solution with the highest fitness value as the current optimal solution; S504, simulating the exploration behavior of the starfish, and updating each solution using the first update formula according to the current optimal solution. The first update formula is: ,in, represents the jth solution in the k+1th iteration, represents the jth solution in the kth iteration, k represents the current iteration number, and j represents the index of the solution. and is a random number between 0 and 1. represents the current optimal solution, represents a random angle; S505, simulating the regeneration stage of the starfish, using the second update formula to update each solution, the second update formula is: , where K represents the maximum number of iterations; S506, when the current number of iterations reaches the maximum number of iterations, the solution with the maximum current fitness value is used as the optimal control parameter set, otherwise return to S503.

6. A fleet internal 5G local area network and satellite communication system according to claim 5, characterized in that: The objective function of the improved starfish optimization algorithm is: ； Among them, F represents the fitness value of the solution, b represents the ship index, represents the communication quality score of the b-th ship.

Images