Method, device and system for dynamically allocating frequency spectrum and electronic equipment

Abstract

The embodiment of the invention discloses a dynamic spectrum allocation method, device and system and electronic equipment, and the method is applied to a control node, and comprises the steps: obtaining an initial scheme of spectrum allocation, the initial scheme is generated based on the data, collected in real time, of target equipment, and the initial scheme comprises at least one spectrum range; dividing the spectral range according to a preset step length to obtain a plurality of candidate schemes; determining an optimal scheme in the plurality of candidate schemes based on the spectrum residual rate, the communication quality index and the load balancing index of the base station; allocating frequency spectrums to the base stations according to the optimal scheme; wherein the target device is one or more devices within the coverage range of the base station.

Description

Technical Field

[0001] This application relates to the field of wireless technology, and in particular to a method, apparatus, system and electronic device for dynamically allocating spectrum. Background Technology

[0002] In maritime communication systems, distributed mobile network node clusters can be formed by deploying unmanned vessels equipped with communication base station equipment. Relying on autonomous navigation and wireless communication technologies, these clusters can achieve wide-area communication coverage in the ocean, emergency communication support, and data backhaul in offshore scenarios. They possess advantages such as high mobility, flexible deployment, and adaptability to complex sea conditions, supporting communication needs in scenarios such as marine monitoring, maritime operations, and disaster relief. They are a crucial mobile infrastructure for marine information technology development. Currently, maritime communication systems focus on technologies such as trajectory prediction and signal transmission optimization, but do not involve spectrum resource allocation technologies. Summary of the Invention

[0003] The purpose of this application is to provide a method, apparatus, system, and electronic device for dynamically allocating spectrum, in order to solve the problems in the prior art.

[0004] To solve the above-mentioned technical problems, the embodiments of this application are implemented as follows: In a first aspect, embodiments of this application provide a method for dynamically allocating spectrum, applied to a control node, the method comprising: An initial scheme for spectrum allocation is obtained, the initial scheme being generated based on data from the target device acquired in real time, and the initial scheme including at least one spectrum range; The spectrum range is divided into multiple candidate schemes according to a preset step size; Based on the base station's spectrum availability, communication quality indicators, and load balancing indicators, the optimal solution is determined from the multiple candidate solutions. Allocate spectrum to the base station according to the optimal scheme described above; The target device refers to one or more devices within the coverage area of ​​the base station.

[0005] Secondly, embodiments of this application provide a device for dynamically allocating spectrum, applied to a control node, the device comprising: An acquisition module is used to acquire an initial scheme for spectrum allocation, the initial scheme being generated based on data collected in real time from the target device, and the initial scheme including at least one spectrum range; The partitioning module is used to partition the spectrum range according to a preset step size to obtain multiple candidate schemes; The determination module is used to determine the optimal solution among the multiple candidate solutions based on the base station's spectrum remaining rate, communication quality indicators, and load balancing indicators. An allocation module is used to allocate spectrum to the base station according to the optimal scheme; The target device refers to one or more devices within the coverage area of ​​the base station.

[0006] Thirdly, embodiments of this application provide a system for dynamically allocating spectrum, the system comprising: Edge nodes are used to collect data from target devices in real time, and to extract features from the data to obtain feature information, which includes at least one of frequency domain features, time features, and spatial features. A central node is used to generate an initial spectrum allocation scheme based on preset weights using a greedy algorithm according to the feature information obtained from the edge nodes. The initial scheme includes at least one spectrum range. A control node is used to obtain the initial scheme generated by the central node, divide the spectrum range into multiple candidate schemes according to a preset step size, determine the optimal scheme among the multiple candidate schemes based on the base station's spectrum remaining rate, communication quality indicators and load balancing indicators, and allocate spectrum to the base station according to the optimal scheme. The target device refers to one or more devices within the coverage area of ​​the base station.

[0007] Fourthly, embodiments of this application provide an electronic device, including a processor, a memory, and a computer program stored in the memory and executable on the processor, wherein the computer program, when executed by the processor, implements the steps of the above-described method for dynamically allocating spectrum.

[0008] Fifthly, embodiments of this application provide a computer-readable storage medium storing a computer program that, when executed by a processor, implements the steps of the above-described method for dynamically allocating spectrum.

[0009] Sixthly, embodiments of this application provide a computer program product, including a computer program that, when executed by a processor, implements the steps of the above-described method for dynamically allocating spectrum.

[0010] As can be seen from the technical solutions provided by the above embodiments of this application, the embodiments of this application obtain an initial scheme for spectrum allocation. This initial scheme is generated based on the data of the target device collected in real time and includes at least one spectrum range. The spectrum range is divided into multiple candidate schemes according to a preset step size. Based on the spectrum remaining rate of the base station, communication quality indicators and load balancing indicators, the optimal scheme is determined among the multiple candidate schemes. The spectrum is allocated to the base station according to the optimal scheme, thus providing a scheme for dynamic spectrum allocation, which can improve the utilization rate of spectrum. Attached Figure Description

[0011] To more clearly illustrate the technical solutions in the embodiments of this application or the prior art, the drawings used in the description of the embodiments or the prior art will be briefly introduced below. Obviously, the drawings described below are only some embodiments recorded in this application. For those skilled in the art, other drawings can be obtained based on these drawings without creative effort.

[0012] Figure 1 A flowchart illustrating a method for dynamically allocating spectrum as provided in an embodiment of this application; Figure 2 A flowchart illustrating the simulated annealing algorithm provided in an embodiment of this application; Figure 3 A flowchart illustrating the process of determining the current optimal solution of the objective function, provided in an embodiment of this application; Figure 4 A schematic diagram illustrating the process of evaluating communication quality provided in an embodiment of this application; Figure 5 A schematic diagram illustrating the process of evaluating spectrum utilization provided in this application embodiment; Figure 6 A schematic diagram of a device for dynamically allocating spectrum provided in an embodiment of this application; Figure 7 A schematic diagram of the structure of a system for dynamically allocating spectrum provided in this application embodiment; Figure 8 This is a schematic diagram of the structure of an electronic device provided in an embodiment of this application. Detailed Implementation

[0013] This application provides a method, apparatus, system, and electronic device for dynamically allocating spectrum.

[0014] To enable those skilled in the art to better understand the technical solutions in this application, the technical solutions in the embodiments of this application will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some embodiments of this application, and not all embodiments. Based on the embodiments of this application, all other embodiments obtained by those skilled in the art without creative effort should fall within the scope of protection of this application.

[0015] The method and apparatus for dynamically allocating spectrum provided in this application are applied to control nodes, specifically in electronic devices, which can be servers. The server can be a standalone server, a server cluster consisting of multiple servers, or a cloud server capable of cloud computing; the specific application is not limited. It should be noted that the server can be deployed on land or in a satellite relay node; the specific deployment is not limited.

[0016] The system for dynamically allocating spectrum provided in this application includes edge nodes, central nodes, and control nodes. The edge nodes and central nodes can be deployed separately on mobile devices, or they can be combined on a single mobile device. Mobile devices include, but are not limited to, ships, vehicles, or robots. For example, the edge nodes and central nodes can be deployed on two unmanned surface vessels, while the control node can be deployed on a land-based server. In practical applications, the central node can be a high-performance computing server (Intel Xeon Platinum + GPU accelerator card) or a cluster collaborative communication module (supporting satellite relay), etc.

[0017] In the above system, there can be one or more edge nodes, one or more central nodes, and one or more control nodes; the specific number is not limited. This distributed architecture constructs a three-level spectrum allocation execution system through edge nodes, central nodes, and control nodes, enabling automated optimization of spectrum allocation.

[0018] The method, apparatus, and system for dynamically allocating spectrum provided in this application are used to dynamically allocate spectrum for base stations. The base station can be deployed on a single device with a central node, or it can be deployed on a single device with an edge node, or the base station can be deployed on a single device with both edge nodes and a central node; the specific deployment is not limited. In practical applications, spectrum can be allocated to one or more base stations, where each base station can include one or more cells, and spectrum can be allocated to one or more cells within each base station. The above describes the dynamic spectrum allocation for base stations, which can be implemented in various scenarios such as land, sea, grassland, or desert.

[0019] The method, apparatus, and system for dynamically allocating spectrum provided in this application can be applied to various scenarios, including but not limited to at least one of the following: Scenario 1: Spectrum allocation and management for daily users or device groups For example, within a coverage area with the same weight, there are ships A and B. Ship A users have lower traffic usage during a specific time period, while ship B users have higher usage. Without adjusting the weights, the coverage resources for ship A can be reduced, such as by dynamically decreasing the corresponding cell carrier frequency, while the coverage resources for ship B can be increased.

[0020] For example, if a ship cluster is included in a coverage area with the same weight, and a new ship enters the coverage area, the spectrum allocation resources of each ship cluster will be reallocated without adjusting the weight.

[0021] For example, when the number of vessels in the coverage area is stable, and there is a sudden increase in business traffic or the addition of other businesses, the usage of vessel and equipment resources in the area can be analyzed, and the allocation of vessel or equipment resources with low business volume can be reduced or reserved resources can be activated.

[0022] For example, coverage areas with different weights but similar service types include vessels C and D. Vessel C has a high weight, but most of its traffic consists of non-real-time, low-priority services, while vessel D has a low weight but experiences bursts of real-time, high-priority services. Without changing the overall weighting framework, the spectrum resources for some low-priority services on vessel C are appropriately reduced, while the spectrum resources for vessel D are correspondingly increased to ensure the timeliness of high-priority services.

[0023] For example, when a vessel's equipment fails and is repaired within the coverage area and reconnects to the network, the system automatically and quickly reallocates appropriate spectrum resources based on the current overall resource usage and the vessel's historical service traffic patterns to ensure the restoration and stable operation of the vessel's services.

[0024] Scenario 2: Emergency Response Scenarios For example, for typical scenarios such as maritime rescue and ocean monitoring, a dynamic priority adjustment and spectrum preemption mechanism can be set up to meet the real-time and reliability requirements of maritime communications. Specifically, 5% to 10% of the spectrum can be reserved as an emergency channel, and the bandwidth of low-priority services such as entertainment data can be dynamically compressed to ensure the highly reliable transmission of rescue commands. This can be achieved by implementing tiered preemption logic based on the vessel's identity tag, such as rescue vessel or work vessel.

[0025] For example, in the scenario of spectrum interconnection during multi-regional emergency response, when multiple maritime areas experience emergency events simultaneously, central nodes in different areas communicate with each other, connecting the spectrum resource pools of each area to achieve cross-regional spectrum redistribution in order to meet large-scale emergency communication needs.

[0026] For example, in emergency service splitting and multi-band adaptation scenarios, for particularly large-scale emergency services such as large-scale maritime military exercises, they are split into multiple sub-services. Based on the characteristics and spectrum requirements of different sub-services, different sub-services are adapted to different frequency bands to avoid excessive congestion on a single frequency band.

[0027] For example, in a grassland fire rescue scenario, the equipment frequency is reduced due to the high temperature at the fire site, the redundant bandwidth is automatically increased to 40%, and the spectrum saturation threshold of the dynamic strategy is automatically reduced from 95% to 60%.

[0028] For example, in a desert sandstorm early warning scenario, the channel capacity is reduced due to the sand particle adsorption effect, and the spectrum saturation threshold of the dynamic strategy is automatically reduced from 80% to 70%.

[0029] Scenario 3: Maritime communication during holidays or special events For example, in predictive spectrum pre-allocation scenarios, communication traffic data from similar periods and events in the past is analyzed and predicted before holidays or special events to plan corresponding spectrum resource allocation schemes in advance. Specifically, before a maritime music festival, sufficient spectrum resources are reserved for audio and video live streaming and audience interaction services to ensure smooth communication during the event.

[0030] For example, in scenarios involving dynamic balancing of multiple service types, special events often see a surge in demand from various service types simultaneously, such as live streaming, real-time data transmission, and social interaction. Real-time monitoring of traffic peaks and troughs for each service type is crucial, allowing for dynamic adjustments to the spectrum resource allocation ratio among different services to ensure that all critical services receive adequate support.

[0031] For example, in cross-regional frequency band sharing optimization scenarios, if similar special events are held simultaneously in multiple maritime areas, different areas can share some idle frequency band resources. Through dynamic coordination between regions, the overall utilization efficiency of spectrum resources can be improved.

[0032] Scenario 4: Communication assurance in harsh environments For example, in spectrum switching scenarios under strong interference, when there is strong electromagnetic interference at sea or the weather is bad and the radio wave propagation environment deteriorates, the system automatically monitors the communication quality of the interfered frequency band. When the communication quality drops to a certain threshold, the system quickly switches the service to the backup low-interference frequency band to ensure communication stability.

[0033] For example, in scenarios where equipment performance adaptively adjusts, harsh environments may affect the communication performance of ships or equipment. Based on performance parameters fed back by the equipment, such as decreased transmit power or reduced receive sensitivity, the spectrum resources allocated to the equipment are dynamically adjusted. For equipment with degraded performance, spectrum resources are appropriately increased to compensate for the communication losses caused by insufficient performance.

[0034] For example, in distributed spectrum optimization scenarios, distributed spectrum optimization strategies are employed in large, harsh environmental coverage areas. Individual vessels or devices collaborate with each other, using ad hoc wireless networking technology to autonomously adjust their spectrum usage strategies based on their location and interference conditions, and share spectrum information with surrounding devices to achieve optimal communication throughout the area.

[0035] Scenario 5: Monitoring Scenario For example, in the scenario of networking scientific research equipment in uninhabited areas, the foundation is unstable due to the melting of permafrost and maintenance bandwidth needs to be reserved. The spectrum saturation threshold of the dynamic strategy is automatically reduced from 85% to 75%.

[0036] For example, in the grassland ecological monitoring network scenario, the base station spacing parameter can be adjusted to 8-12km. Considering vegetation penetration optimization, the operating frequency band can be adjusted to prioritize 700MHz-2.6GHz. The dynamic strategy automatically increases the weight of fire warning period to 75%.

[0037] For example, in desert oil and gas field monitoring scenarios, energy efficiency optimization can switch to the 915MHz ISM band at night, reducing power consumption by 38%.

[0038] For example, in uninhabited scientific research scenarios, the simplified transmission protocol dynamically adjusts and compresses the packet header to 12 bytes.

[0039] It should be noted that if multiple scenarios mentioned above occur simultaneously, the reordering mechanism will invoke a strategy to reallocate resources. Through the synergistic effect of the above mechanisms, the system can complete spectrum resource reordering within 30 seconds, achieving priority preemption and continuous guarantee for emergency communications. Compared with traditional algorithms, the response speed is improved by 80%, and the success rate of critical communications is increased to over 99%.

[0040] like Figure 1 As shown in the figure, this application provides a method for dynamically allocating spectrum, applied to a control node, which may include the following steps: S102: Obtain an initial scheme for spectrum allocation, the initial scheme being generated based on data from the target device acquired in real time, and the initial scheme including at least one spectrum range.

[0041] The target device refers to one or more devices within the coverage area of ​​the base station. For example, the target device includes, but is not limited to, ships, vehicles, IoT devices, or robots.

[0042] The initial spectrum allocation scheme can be obtained using a greedy algorithm based on preset weights. For example, the greedy algorithm can be applied to a Field Programmable Gate Array (FPGA). FPGA acceleration enables rapid weight updates, such as 1000 weight updates per second, thereby improving algorithm efficiency.

[0043] S104: Divide the spectrum range according to a preset step size to obtain multiple candidate schemes.

[0044] For example, the spectrum range is 5MHz, corresponding to the frequency band With a preset step size of 0.1MHz, dividing the frequency band according to the preset step size yields the following results: Different frequency band combination schemes within the range.

[0045] S106: Based on the base station's spectrum remaining rate, communication quality indicators, and load balancing indicators, determine the optimal solution from multiple candidate solutions.

[0046] In this embodiment, simulated annealing can be used to determine the optimal solution. The simulated annealing algorithm can be applied to a graphics processing unit (GPU), accelerating the solution of the objective function through parallel computing. Iterative convergence to a near-global optimum within the target number of iterations improves algorithm efficiency and adapts to wide-area coverage requirements. The simulated annealing algorithm also offers advantages such as globally optimizing spectrum fragmentation and enabling cross-base station resource sharing.

[0047] S108: Allocate spectrum to the base station according to the optimal scheme.

[0048] See Figure 2 In this embodiment of the application, step S106 may include: S202: Calculate the objective function of candidate solutions based on the base station's spectrum surplus rate, communication quality indicators, and load balancing indicators.

[0049] In this embodiment of the application, step S202 may include: Based on the base station's spectrum availability, communication quality metrics, and load balancing metrics, the objective function of the candidate solutions is calculated using the following formula: ; ; ; ; ; Where f represents the objective function of the candidate scheme, R represents the spectrum surplus rate, Q represents the communication quality index, and B represents the load balancing index. Indicates the total bandwidth of the available spectrum. Indicates the allocated spectrum bandwidth. This represents the average bit error rate. The expression represents the average transmission delay, 'a' represents the bit error rate threshold, 'b' represents the transmission delay threshold, and 'n' represents the number of base stations. This indicates the load of each base station. This indicates the average load of the base station. The weights representing the spectrum surplus rate The weights of communication quality metrics are indicated. This indicates the weight of the load balancing metric.

[0050] Here, 'a' and 'b' can be set as needed, such as setting 'a='. Setting b=100, etc., is not limited to any specific value. , and You can also set it as needed, such as setting =0.4, =0.4, =0.2 etc.

[0051] S204: Calculate the objective function difference based on the objective function of the candidate solutions, calculate the acceptance probability based on the current temperature and the objective function difference, generate random numbers and compare them with the acceptance probability to determine the current optimal solution of the objective function.

[0052] S206: Based on the current temperature and cooling rate, repeatedly iterate to determine the current optimal solution until the current temperature drops to the target temperature or the number of iterations reaches the target number, and obtain the global optimal solution of the objective function as the optimal scheme for spectrum allocation.

[0053] The cooling rate is used to reduce the current temperature. In each iteration, T = T × α. As the temperature gradually decreases, the search range of the algorithm gradually shrinks and the search accuracy gradually improves.

[0054] The current temperature can be set as the initial temperature during initialization, and then gradually reduced according to the cooling rate. The initial temperature, cooling rate, target temperature, and target number of cycles can all be preset as needed; for example, the initial temperature can be set to 100, and the cooling rate can be set to... The target temperature can be set to a value less than 1, such as 0.01, and the target number of times can be set to 1000, etc.

[0055] In this embodiment, a higher initial temperature enables the algorithm to search the solution space more extensively in the initial stage, which helps to discover potential high-quality spectrum allocation schemes; while the cooling rate This allows for precise control of the rate at which the temperature decreases. A suitable cooling rate ensures that the algorithm gradually converges to a better solution during the search process.

[0056] See Figure 3 In this embodiment of the application, step S204 may include: S302: Determine the current candidate solution and alternative candidate solutions from multiple candidate solutions.

[0057] S304: Calculate the difference in objective functions based on the objective functions of the current candidate solution and the alternative candidate solutions.

[0058] S306: Calculate the acceptance probability P based on the difference between the current temperature and the objective function.

[0059] S308: Generate a random number r, where 0 ≤ r ≤ 1.

[0060] The random number r mentioned above can be generated using a random number generator.

[0061] S310: If r≤P, take the candidate solution as the current optimal solution of the objective function (even if the current candidate solution is better than the candidate solution). If r>P, take the current candidate solution as the current optimal solution of the objective function (reject the candidate solution and keep the current candidate solution).

[0062] In this embodiment of the application, the formula for calculating the acceptance probability P can be as follows: ; Where P represents the probability of acceptance. Let T represent the output of the objective function, and let T represent the current temperature.

[0063] For example, the difference of the objective function Current temperature ,but This indicates that, under these circumstances, there is a high probability of accepting the alternative candidate.

[0064] The acceptance probability P determines whether to accept the current candidate solution (even a poor one). Combining this with random numbers to accept candidate solutions with a certain probability allows for the avoidance of local optima, broadens the search space, and ultimately outputs a globally optimized spectrum allocation scheme. This probability-based acceptance mechanism balances the algorithm's exploration of new solutions with maintaining the current relatively good solution, ensuring that potential global optima are not missed and guaranteeing the algorithm's stability and convergence to a certain extent, thereby improving spectrum utilization and communication performance in the communication system.

[0065] In this embodiment of the application, the above method may further include: in the event of an emergency, adjusting the optimal solution according to at least one of the following: 1) Temporarily allocate idle frequency bands and / or reserve frequency bands for emergency services to ensure bandwidth availability for sudden demand.

[0066] The reserved frequency bands can be set as needed, such as reserving 20% ​​of the global spectrum resources as reserved frequency bands. Moreover, the reserved frequency bands can be adjusted according to changes in the scenario, such as adjusting the reserved frequency bands during peak communication periods such as typhoon relief or transoceanic shipping routes.

[0067] Among them, the temporary allocation of idle frequency bands and / or the reservation of frequency bands for emergency services can be set for a period of no more than a certain duration, such as no more than 30 minutes, so as to avoid long-term occupation and affect normal services.

[0068] Among these methods, deep neural networks can be used to predict future hourly spectrum demand based on historical and real-time data, thereby reserving suitable frequency bands.

[0069] 2) Increase the weight of emergency business.

[0070] If an SOS signal is received from the Maritime Safety Administration, the weight of emergency services will be increased from the initial value of 5 to 8, or from 8 to 10, to ensure that emergency communication needs have an absolute advantage in spectrum allocation decisions.

[0071] 3) Temporarily relax communication quality requirements for non-emergency services.

[0072] For example, the bit error rate threshold is from 10 -5 Relaxed to 10 -4 More computing resources are allocated to signal optimization for emergency communications, such as using gradient descent to adjust transmit power and Kalman filtering to suppress noise, to ensure that emergency services have an SS-RSRP ≥ -98dBm and an SS-SINR ≥ -3dB.

[0073] In this embodiment, emergency services refer to services with high communication demands and strict timeliness requirements, such as disaster relief services. Non-emergency services refer to ordinary services, i.e., services used for normal communication, such as voice services, video services, or internet access services.

[0074] In this embodiment of the application, the above method may further include at least one of the following: 1) If the load on a certain frequency band in the optimal solution exceeds the load threshold, reduce the weight of ordinary services in the frequency band and reclaim idle frequency bands or frequency bands occupied by low priority.

[0075] For example, if the current load of a base station exceeds 80% of its maximum load, the weight of ordinary services in that frequency band will be reduced by 30%.

[0076] In this embodiment, a frequency band is considered an idle band when the signal power is below the interference threshold. Idle frequency bands can be monitored in real time using a combination of energy detection and matched filter detection techniques.

[0077] 2) Adjust the reserved global frequency bands during peak communication periods or important business operations.

[0078] Among them, important business refers to business with higher priority and greater importance than ordinary business.

[0079] 3) When there are multiple base stations and their coverage areas overlap, the spectrum in the optimal scheme is shared by multiple base stations.

[0080] In this scenario where multiple base stations share the optimal spectrum resources, spectrum reuse can be achieved, global resource scheduling can be implemented for overlapping coverage areas, and cross-cluster interference can be avoided.

[0081] See Figure 4 In this embodiment of the application, the above method can also evaluate communication quality, specifically including the following steps: S402: Real-time monitoring of Synchronization Signal-Reference Signal Received Power (SS-RSRP), Synchronization Signal-Signal to Interference plus Noise Ratio (SS-SINR), and Bit Error Rate (BER).

[0082] In this embodiment of the application, the signal-to-interference-plus-noise ratio (SIR) of the synchronization signal can be calculated using the following formula: ; in, The signal-to-interference-plus-noise ratio (SIR) of the synchronization signal is expressed in dB. This indicates the average power of the signal (unit: W). This represents the average power of the noise (unit: W).

[0083] In this embodiment of the application, the bit error rate can be calculated using the following formula: ; Here, BER represents the bit error rate. This indicates the number of bits that were corrupted during transmission. Indicates the total number of bits transmitted.

[0084] S404: If the SS-RSRP is lower than the first threshold, the dynamic power adjustment algorithm is used to optimize the transmitter parameters.

[0085] The first threshold mentioned above can be preset as needed, and the specific value is not limited.

[0086] For example, if the first threshold is -98dBm, and SS-RSRP is detected to be less than -98dBm, a dynamic power adjustment algorithm (such as gradient descent) is triggered to optimize the transmitter parameters, such as adjusting the transmit power and optimizing the antenna direction, in order to improve the SS-RSRP strength and ensure communication quality.

[0087] S406: If the SS-SINR is lower than the second threshold, then use adaptive filtering techniques to suppress noise.

[0088] The aforementioned second threshold can be preset as needed, and the specific value is not limited.

[0089] For example, if the second threshold is -3dB, and SS-SINR < -3dB is detected, adaptive filtering techniques (such as Kalman filtering) are triggered to suppress noise. Measures such as adjusting the signal modulation method and increasing signal coding redundancy are used to improve SS-SINR and ensure the accuracy and anti-interference capability of communication.

[0090] S408: If the BER is higher than the third threshold, forward error correction coding is used to correct it or retransmission is performed.

[0091] The aforementioned third threshold can be preset as needed, and the specific value is not limited.

[0092] For example, the third threshold is set to 10. -5 If BER>10 is detected -5 If the error is detected, the error correction mechanism is activated. Using forward error correction coding such as LDPC error correction coding, the bit error rate can be corrected within 1 second, thereby reducing the bit error rate and ensuring the correctness of data transmission.

[0093] The mechanism described above, which adjusts based on the first, second, and third thresholds, can ensure the stability and reliability of communication.

[0094] See Figure 5 In this embodiment of the application, the above method can also evaluate the spectrum utilization rate, which may specifically include the following steps: S502: Calculate the spectrum hole rate based on the bandwidth idle ratio threshold.

[0095] In this embodiment, the spectrum hole rate is used to measure the degree to which spectrum resources are not effectively utilized. Spectrum switching can be caused by various factors, such as changes in communication demand and signal interference.

[0096] Step S502 above may include: The spectral void ratio within the statistical period is calculated using the following formula: ; ; ; in, This indicates the duration (in seconds) that the k-th frequency band is occupied within the statistical period. This represents the end time of the i-th use of the k-th frequency band within the statistical period. This represents the start time of the i-th use of the k-th frequency band within the statistical period, and N represents the total number of times the k-th frequency band is used within the statistical period. This represents the idle time of the k-th frequency band within the statistical period. Indicates the statistical period. Indicates the spectral void ratio. Indicates the threshold for continuous idle time (e.g.) ), Indicates the bandwidth idle ratio threshold (e.g.) ), This indicates an exponential function; the exponential function takes the value 1 if the condition within the parentheses is met, and takes the value 0 if the condition within the parentheses is not met. This represents the total bandwidth of the k-th frequency band within the statistical period. This represents the idle bandwidth of the k-th frequency band within the statistical period, where K represents the number of frequency bands.

[0097] S504: Calculate the spectrum switching frequency of the base station.

[0098] In this embodiment of the application, the above-mentioned spectrum switching frequency can be calculated using the following formula: ; in, Indicates the frequency of spectrum switching. This indicates the total number of network handovers within the statistical period. Indicates the statistical period.

[0099] The upper limit threshold for the spectrum switching frequency can be preset, such as set to... ,Right now ≤10 times / hour, if If the frequency exceeds 8 times per hour, a handover suppression algorithm can be activated to reduce the spectrum handover frequency. Controlling the spectrum handover frequency within a reasonable range helps maintain a stable and efficient communication system.

[0100] S506: When the spectrum switching frequency exceeds the switching threshold, use a switching suppression algorithm to reduce the spectrum switching frequency.

[0101] S508: When the base station load exceeds the load threshold, reduce the bandwidth idle ratio threshold to adjust the spectrum hole rate and increase the upper limit threshold of the spectrum switching frequency.

[0102] In this embodiment of the application, the load of the base station can be calculated based on indicators such as the amount of service data processed by the base station and the number of connected users, which will not be elaborated here.

[0103] The load threshold can be preset as needed. For example, the load threshold can be set to 80% of the maximum load. When the base station load L > 0.8Lmax, based on the principle of prioritizing communication continuity under high load, the bandwidth idle ratio threshold can be adjusted. The requirement for spectrum switching frequency is relaxed by reducing it from 30% to 20% and increasing the spectrum switching frequency from 10 times / hour to 12 times / hour, thereby enabling intelligent adjustment of load-sensitive parameters.

[0104] In this embodiment of the application, the above method may further include: The activity level of each frequency band within the statistical period is calculated using the following formula: ; in, This represents the activity level of the k-th frequency band within the statistical period. , This indicates the number of times the k-th frequency band is used within the statistical period. This represents the bandwidth (in MHz) of the k-th frequency band within the statistical period. Indicates the statistical period; If the activity level is greater than the activity threshold, the k-th frequency band is split into multiple sub-frequency bands for allocation.

[0105] The activity threshold can be set as needed, for example, setting the activity threshold to 15. When the frequency band splitting mechanism is triggered, the kth frequency band can be split into multiple sub-frequency bands for allocation, thus achieving dynamic adjustment.

[0106] In this embodiment, high activity indicates frequent demand for the frequency band, and the existing bandwidth may not be able to flexibly adapt to the bandwidth requirements of different services, easily leading to resource waste or contention. By splitting the frequency band into multiple sub-bands for allocation through a splitting mechanism, each sub-band can be independently allocated to different users or services, thereby matching demand and improving the utilization efficiency of spectrum resources.

[0107] In this embodiment of the application, the above method may further include: Real-time monitoring of the target device's speed; When the target device's moving speed exceeds a speed threshold, reduce the spectrum sensing cycle and reduce the continuous idle time threshold. , Used to calculate the spectral hole rate.

[0108] The speed threshold can be set as needed. For example, setting the speed threshold to 15 knots will shorten the spectrum sensing period from 60 seconds to 20 seconds when the target ship's speed v > 15 knots, due to the rapid changes in channel state caused by high-speed movement. Furthermore, the continuous idle time threshold can also be adjusted. The time was reduced from 300s to 180s to achieve adaptive adjustment of movement parameters.

[0109] In this embodiment of the application, a trigger condition for spectrum allocation can also be set, and the above method will be executed when the trigger condition is met. The trigger condition may include at least one of the following: 1) Regular allocation trigger: Comprehensive utilization index When the saturation threshold is set and the spectrum fragmentation rate is >15%, a greedy algorithm is used to prioritize the integration of fragmented frequency bands to improve available bandwidth.

[0110] 2) Emergency Allocation Forced Trigger: Detection of SOS signals from rescue vessels or a sudden increase in the error rate ( If the frequency remains below 10 seconds, the spectrum preemption strategy will be activated, releasing the reserved 20% emergency bandwidth.

[0111] 3) Adaptive environment triggering: When environmental changes cause a decrease in the confidence of the neural network model of edge nodes (such as the variance of CNN branch output > preset threshold), the model self-calibration mechanism is activated, and the simulated annealing algorithm is triggered to re-optimize the allocation scheme.

[0112] In this embodiment of the application, the above method may further include: The predicted data for spectrum demand are calculated using a predictive model; Obtain actual data on spectrum allocation; Calculate the error between the predicted data and the actual data, assess the accuracy of the prediction model based on the error, and adjust the prediction model accordingly.

[0113] The error between the predicted data and the actual data can be represented by the root mean square error (RMSE), which can be calculated using the following formula: ; Where n represents the number of data samples, This represents the predicted value of the i-th sample. This represents the actual value of the i-th sample. If... If the predicted value is small and the deviation between the predicted and actual values ​​is within an acceptable range, it indicates that the prediction model has high accuracy. Conversely, if the predicted value is large, the prediction model needs to be adjusted and optimized, such as by reselecting model parameters or increasing training data, to improve its prediction accuracy.

[0114] The above process ensures that the prediction model can accurately estimate future spectrum demand. By constructing a dynamic parameter intelligent adjustment mechanism and supplementing it with an accuracy verification step, intelligent spectrum allocation performance evaluation can be achieved, thereby ensuring the environmental adaptability and reliability of the spectrum allocation strategy.

[0115] The actual spectrum allocation data can be obtained by installing high-precision data acquisition equipment on each base station and target device, ensuring the accuracy and integrity of the data.

[0116] This system can simultaneously collect predicted spectrum demand data and actual spectrum allocation data from the prediction model at different time periods and in different regions, aligning the predicted and actual data according to the same time intervals and frequency bands. Specifically, different time periods include different weekdays, weekends, and different seasons; different regions include different nearshore areas, offshore areas, and areas near different shipping routes; and time intervals can be hourly or daily.

[0117] The method provided in this application embodiment obtains an initial spectrum allocation scheme, which is generated based on real-time collected data from the target device and includes at least one spectrum range. The spectrum range is divided into multiple candidate schemes according to a preset step size. Based on the base station's spectrum remaining rate, communication quality indicators, and load balancing indicators, the optimal scheme is determined among the multiple candidate schemes. The spectrum is allocated to the base station according to the optimal scheme, providing a dynamic spectrum allocation scheme that can improve spectrum utilization, reduce spectrum resource waste, and achieve spectrum resource optimization.

[0118] Moreover, it can effectively cope with complex and ever-changing communication environments and meet diverse communication needs. Furthermore, the introduction of spectrum reuse and sharing mechanisms enables rapid response and priority protection of critical communications in emergency situations, reducing the possibility of delays in critical rescue communications and further enhancing the flexibility and reliability of spectrum allocation. Through the combination of greedy algorithms and simulated annealing algorithms, spectrum allocation strategies can be dynamically adjusted to meet the communication needs of different service types and priorities. By monitoring spectrum resources and communication quality in real time, it can quickly respond to environmental changes and ensure communication quality.

[0119] The above describes a method for dynamically allocating spectrum according to embodiments of this application. Based on the same idea, embodiments of this application also provide a device for dynamically allocating spectrum, applied to a control node, such as... Figure 6 As shown, it may include: The acquisition module 601 is used to acquire an initial scheme for spectrum allocation. The initial scheme is generated based on data from the target device acquired in real time, and the initial scheme includes at least one spectrum range.

[0120] The partitioning module 602 is used to partition the spectrum range according to a preset step size to obtain multiple candidate schemes.

[0121] The determination module 603 is used to determine the optimal solution from multiple candidate solutions based on the base station's spectrum remaining rate, communication quality indicators, and load balancing indicators.

[0122] The allocation module 604 is used to allocate spectrum to the base station according to the optimal scheme.

[0123] The target device refers to one or more devices within the coverage area of ​​the base station.

[0124] In this embodiment of the application, the determining module 603 is used for: The objective function of candidate solutions is calculated based on the base station's spectrum availability, communication quality indicators, and load balancing indicators. The objective function difference is calculated based on the objective function of the candidate solutions. The acceptance probability is calculated based on the current temperature and the objective function difference. Random numbers are generated and compared with the acceptance probability to determine the current optimal solution of the objective function. The optimal solution is determined by iteratively analyzing the current temperature and cooling rate until the current temperature drops to the target temperature or the number of iterations reaches the target number. The global optimal solution of the objective function is then obtained as the optimal scheme for spectrum allocation.

[0125] In this embodiment of the application, the determining module 603 calculates the objective function of the candidate scheme based on the base station's spectrum remaining rate, communication quality indicators, and load balancing indicators, including: Based on the base station's spectrum availability, communication quality metrics, and load balancing metrics, the objective function of the candidate solutions is calculated using the following formula: ; ; ; ; ; Where f represents the objective function of the candidate scheme, R represents the spectrum surplus rate, Q represents the communication quality index, and B represents the load balancing index. Indicates the total bandwidth of the available spectrum. Indicates the allocated spectrum bandwidth. This represents the average bit error rate. The expression represents the average transmission delay, 'a' represents the bit error rate threshold, 'b' represents the transmission delay threshold, and 'n' represents the number of base stations. This indicates the load of each base station. This indicates the average load of the base station. The weights representing the spectrum surplus rate The weights of communication quality metrics are indicated. This indicates the weight of the load balancing metric.

[0126] In this embodiment of the application, the determining module 603 calculates the difference in objective functions based on the objective functions of the candidate schemes, calculates the acceptance probability based on the current temperature and the difference in objective functions, generates random numbers and compares them with the acceptance probability, and determines the current optimal solution of the objective function, including: Identify the current candidate solution and alternative candidate solutions from multiple candidate solutions; Calculate the difference in objective functions based on the objective functions of the current candidate solutions and the alternative candidate solutions; Calculate the acceptance probability P based on the difference between the current temperature and the objective function; Generate a random number r. If r ≤ P, then the candidate solution is taken as the current optimal solution of the objective function. If r > P, then the current candidate solution is taken as the current optimal solution of the objective function. Where 0 ≤ r ≤ 1.

[0127] In this embodiment of the application, the above-described apparatus is further used for: In the event of an emergency, the optimal solution shall be adjusted according to at least one of the following: Temporarily allocate idle frequency bands and / or reserve frequency bands for emergency services; Increase the weight given to urgent business; The communication quality requirements for non-emergency services will be temporarily relaxed.

[0128] In this embodiment of the application, the above-described apparatus is further configured to perform at least one of the following: If the load on a certain frequency band in the optimal solution exceeds the load threshold, reduce the weight of ordinary services in the frequency band and reclaim idle frequency bands or frequency bands occupied by low priority. Adjustments will be made to the reserved global frequency bands during peak communication periods or important business operations. When there are multiple base stations and their coverage areas overlap, the spectrum in the optimal scheme is shared among multiple base stations.

[0129] In this embodiment of the application, the above-described apparatus is further used for: Real-time monitoring of synchronization signal reference signal received power (SS-RSRP), synchronization signal signal-to-interference-plus-noise ratio (SS-SINR), and bit error rate; If the SS-RSRP is lower than the first threshold, the dynamic power adjustment algorithm is used to optimize the transmitter parameters; If the SS-SINR is lower than the second threshold, then adaptive filtering techniques are used to suppress noise. If the bit error rate is higher than the third threshold, forward error correction coding is used for correction or retransmission is performed.

[0130] In this embodiment of the application, the above-described apparatus is further used for: Calculate the spectrum hole rate based on the bandwidth idle ratio threshold; Calculate the spectrum switching frequency of the base station; When the spectrum switching frequency exceeds the switching threshold, a switching suppression algorithm is used to reduce the spectrum switching frequency. When the base station load exceeds the load threshold, the bandwidth idle ratio threshold is lowered to adjust the spectrum hole rate, and the upper limit threshold of the spectrum switching frequency is increased.

[0131] In this embodiment of the application, the above-mentioned device calculates the spectral hole rate based on a bandwidth idle ratio threshold, including: The spectral void ratio within the statistical period is calculated using the following formula: ; ; ; in, This represents the duration of time the k-th frequency band is occupied within the statistical period. This represents the end time of the i-th use of the k-th frequency band within the statistical period. This represents the start time of the i-th use of the k-th frequency band within the statistical period, and N represents the total number of times the k-th frequency band is used within the statistical period. This represents the idle time of the k-th frequency band within the statistical period. Indicates the statistical period. Indicates the spectral void ratio. Indicates the threshold for continuous idle time. This indicates the threshold for the proportion of bandwidth that is not in use. This indicates an exponential function; the exponential function takes the value 1 if the condition within the parentheses is met, and takes the value 0 if the condition within the parentheses is not met. This represents the total bandwidth of the k-th frequency band within the statistical period. This represents the idle bandwidth of the k-th frequency band within the statistical period, where K represents the number of frequency bands.

[0132] In this embodiment of the application, the above-described apparatus is further used for: The activity level of each frequency band within the statistical period is calculated using the following formula: ; in, This indicates the activity level of the k-th frequency band within the statistical period. This indicates the number of times the k-th frequency band is used within the statistical period. This represents the bandwidth of the k-th frequency band within the statistical period. Indicates the statistical period; If the activity level is greater than the activity threshold, the k-th frequency band is split into multiple sub-frequency bands for allocation.

[0133] In this embodiment of the application, the above-described apparatus is further used for: Real-time monitoring of the target device's speed; When the target device's moving speed exceeds a speed threshold, reduce the spectrum sensing cycle and reduce the continuous idle time threshold. , Used to calculate the spectral hole rate.

[0134] The apparatus provided in this application embodiment can execute the method provided in any of the above method embodiments. For detailed process, please refer to the description in the method embodiments, which will not be repeated here.

[0135] The apparatus provided in this application obtains an initial spectrum allocation scheme, which is generated based on real-time collected data from the target device and includes at least one spectrum range. The spectrum range is divided into multiple candidate schemes according to a preset step size. Based on the base station's spectrum remaining rate, communication quality indicators, and load balancing indicators, the optimal scheme is determined among the multiple candidate schemes. The spectrum is allocated to the base station according to the optimal scheme, thus providing a dynamic spectrum allocation scheme that can improve spectrum utilization.

[0136] See Figure 7 This application also provides a system for dynamically allocating spectrum, which may include: Edge node 701 is used to collect data from the target device in real time, and to extract feature information from the collected data. The feature information includes at least one of frequency domain features, time features, and spatial features.

[0137] The central node 702 is used to generate an initial spectrum allocation scheme based on preset weights using a greedy algorithm according to the feature information obtained from the edge nodes. The initial scheme includes at least one spectrum range.

[0138] Control node 703 is used to obtain the initial scheme generated by the central node, divide the spectrum range according to a preset step size to obtain multiple candidate schemes, and determine the optimal scheme among multiple candidate schemes based on the base station's spectrum remaining rate, communication quality indicators and load balancing indicators, and allocate spectrum to the base station according to the optimal scheme.

[0139] The target device refers to one or more devices within the coverage area of ​​the base station.

[0140] Among them, control node 703 and Figure 6 The devices shown have the same function and can achieve the same technical effect, so they will not be described in detail here.

[0141] In this embodiment of the application, the edge node 701 extracts feature information from the collected data, including: If the proportion of missing data in the collected data is less than or equal to the preset proportion, the missing data will be filled with the mean or median of the non-missing data. If the proportion of missing data in the collected data is greater than the preset proportion, use a time series model or machine learning regression model to predict and fill in the missing data. Remove noisy data from the filled data; Feature information is obtained by extracting features from the data after noise removal. Each parameter in the feature information is labeled based on its corresponding initial threshold, and the label is marked as excellent or poor.

[0142] Edge node 701 can perform a comprehensive scan of the collected data. If the proportion of missing data in a specified frequency band or time series is less than 20%, the mean or median of the existing data in that frequency band or time series is calculated to fill in the missing data. If the proportion of missing data in the collected data is greater than 20% and the collected data has time series characteristics or correlation, a time series model (such as the ARIMA model) or a machine learning regression model (such as linear regression, decision tree regression, etc.) is selected to predict and fill in the missing data.

[0143] Among them, edge node 701 can choose a method that combines mean filtering and three times the standard deviation to remove data that deviates from the mean by more than three times the standard deviation, so as to remove noisy data in the filled data.

[0144] The feature information extracted by edge node 701 includes at least one of frequency domain features, temporal features, and spatial features. For example, the feature information includes features such as temporal distribution, frequency band usage frequency, and the relative position of the target device.

[0145] Among them, edge node 701 can mark the initial threshold for spectrum saturation as 80% of the total spectrum bandwidth. This initial threshold can be regarded as the optimal balance point between channel capacity and communication quality, thus reserving 20% ​​bandwidth to cope with sudden demand. Furthermore, the bit error rate can also be marked: below... Marked as excellent, higher than The following are marked as poor. Transmission delay is also marked: less than 150ms is excellent, and greater than 200ms is poor.

[0146] Among them, the edge node 701 can collect BeiDou positioning data of the target device at a frequency of 100Hz, and obtain the target device's location information such as latitude and longitude (accurate to 6 decimal places), direction of movement (accurate to 0.1 degrees), speed (accurate to 0.1 knots), base station location, communication traffic (byte level), service type (voice / video / data), priority (high / medium / low) and other location information data.

[0147] In this embodiment, the target device can be a fixed 5G IoT device on a private network, and the edge node 701 can collect data such as that from marine ranches or offshore wind power generation. Specifically, information such as the data volume (accurate to bytes), service type (voice / video / data), and priority (high / medium / low) of the IoT device can be collected once per second.

[0148] In this embodiment, edge node 701 can also scan the 0-6GHz mid-frequency band and the 24-30GHz high-frequency band in real time using energy detection and matched filtering technology, updating the spectrum hole map (resolution 0.1MHz) every 10ms, and outputting indicators such as SS-RSRP and SS-SINR. Mean filtering and a three-standard-deviation method are used to denoise the original data, and features such as time distribution, frequency band utilization, and relative distance between ships are extracted. The data is then transmitted to central node 702 via a 5G NR-U wireless link.

[0149] In this application embodiment, the parameters related to the marine environment include, but are not limited to: salt spray, wave height, seabed, wind speed, wind direction, rainfall, seawater temperature, seawater pH, distribution of marine life, number, location, and relative movement speed of target vessels or IoT devices.

[0150] In this embodiment, the edge node 701 can also combine sensor technology, satellite remote sensing technology, etc. to collect data, ensuring the timeliness and accuracy of data collection.

[0151] When severe weather causes unstable data transmission signals from IoT devices, in addition to changing the frequency band, data caching and burst transmission can also be used. Specifically, after temporarily caching the collected data, it can be transmitted back to the central node 702 in a burst when the weather conditions improve or the signal strengthens, thereby reducing the risk of data loss and ensuring data integrity.

[0152] For information on target vessels or IoT devices, their number, location, and relative speed can be obtained using the BeiDou satellite positioning system and radar detection technology. For salt spray information, high-precision salt spray sensors can be used to monitor salt spray concentration and distribution in real time.

[0153] Regarding seabed information, if a shipborne base station encounters situations where it cannot reach its planned route due to reefs or other obstacles, the central node 702 can analyze the situation and adjust the shipborne base station's navigation path to avoid these special sea areas as much as possible. Alternatively, it can flexibly change frequency band usage strategies based on real-time signal strength and coverage requirements. For example, it can switch from high-frequency bands to low-frequency bands or from narrow-beam coverage to wide-beam coverage when necessary to improve coverage distance and quality, ensuring stable data transmission.

[0154] When a shipborne base station encounters special sea areas such as restricted navigation zones or areas with strong currents, the central node 702 can use a pre-built multi-objective optimization algorithm model to not only consider the strategy of changing frequency band usage, but also take into account factors such as the shipborne base station's moving speed and turning angle, so as to achieve coordinated optimization of multiple strategies.

[0155] Collecting parameters such as seawater temperature, pH, and marine life distribution provides a comprehensive understanding of the marine environment and richer data support for relevant decision-making. For example, changes in seawater temperature and pH can affect marine ecosystems and the navigation safety of target vessels; collecting this data can provide more reference strategies for marine ecological protection and navigation planning.

[0156] In this embodiment, the edge node 701 can integrate a BeiDou positioning module (accuracy up to 0.1 meters), a spectrum sensing sensor (supporting 24GHz+ millimeter wave band), and an edge computing unit (equipped with NVIDIA Jetson Xavier NX).

[0157] In this embodiment, the edge node 701 can use a neural network model to extract features from the data to obtain feature information. The neural network model includes: Convolutional Neural Network (CNN) layers are used to extract features from input data and output frequency domain features; Pooling layers are used to transform frequency domain features; The Long Short-Term Memory Network (LSTM) layer is used to extract features from the input data and output temporal and spatial features. The first fully connected layer is used to perform multi-dimensional semantic parsing of frequency domain features, time domain features and spatial features according to multiple preset semantic subspaces, and learn feature associations; The second fully connected layer is used to compress the feature dimension of the output of the first fully connected layer.

[0158] The CNN layer is responsible for capturing local features, while the pooling layer enhances the robustness of the data to transformations such as translation and rotation. The LSTM layer implements the functionality of a Recurrent Neural Network (RNN). The fully connected layer classifies and predicts spectral features, ultimately outputting the analysis results.

[0159] The aforementioned fusion of CNN and RNN constitutes a Dynamic Spatial-Temporal Fusion Network (DSTFN) architecture, which enables effective analysis and processing of spectral features, providing support for subsequent spectrum resource management. Customized designs are implemented to address the multi-dimensional spectral data characteristics (including temporal fluctuations, frequency distribution, and spatial location correlation) of various communication scenarios, overcoming the limitations of traditional single-network structures in feature extraction.

[0160] In this structure, CNN and LSTM form a parallel branching structure, and feature path dynamic selection is achieved through an adaptive gate mechanism.

[0161] The CNN layer can employ an improved dense convolutional block, cascading 16 3×3 convolutional kernels (stride 1, padding 1) to form a local feature extraction chain. Combined with boundary-preserving pooling, the spatial boundary information of the spectral data is preserved through edge interpolation during the 2×2 pooling process, solving the problem of blurred feature edges caused by traditional max pooling and improving the frequency band boundary recognition accuracy by about 20%.

[0162] The LSTM layer includes a multi-scale time window attention module, which processes spectral sequences with three time granularities of 10 minutes, 30 minutes, and 1 hour in parallel in the 128-dimensional hidden layer. It automatically focuses on sudden spectral change events (such as frequency band changes caused by typhoon interference) through attention weights. Compared with the fixed time window model, the accuracy of capturing sudden features is improved by about 5%.

[0163] In addition, a cross-modality interaction unit can be introduced before the fully connected layer to achieve nonlinear fusion of the frequency domain features output by the CNN layer with the temporal and spatial features output by the LSTM layer through an adaptive weight matrix.

[0164] The first fully connected layer (256 dimensions) employs a grouped convolution method to divide the features into multiple semantic subspaces. These include subspaces for low- and mid-frequency band utilization, communication service type, signal strength distribution, transmission delay features, base station load balancing, spectrum hole distribution, and sudden interference events. These subspaces independently learn feature association patterns through grouped convolution, enabling multi-dimensional semantic parsing of the spectrum data and improving the model's ability to classify and predict spectrum features under complex sea conditions.

[0165] The second fully connected layer (128 dimensions) introduces low-rank regularization constraints, compressing the feature dimension through matrix singular value decomposition. When the feature dimension is compressed by 5% after matrix singular value decomposition, the data integrity can be considered high, and the computational cost can be reduced by 20% while retaining 95% of the key information. Although this method loses 5% of the key information, it still reduces the computational cost by 20%, which is highly valuable in situations where energy is extremely scarce at maritime base stations, meeting the lightweight deployment requirements of unmanned shipborne equipment.

[0166] In the embodiment of the present application, the edge node 701 can also perform self-calibration using an environment perception model, taking environmental parameters such as the real-time ship speed (v) and sea wave level (s) obtained by Beidou positioning as conditional inputs, and dynamically adjusting network parameters in the following ways: When it is detected that v > 15 knots or s > 4 levels, the fast feature update mode is triggered, the CNN convolution step size is adaptively adjusted to 2, and the weight of the LSTM forget gate is increased by 0.2 to improve the response speed to the rapid spectrum changes in the high-speed moving scenario; Based on the online parameter optimization algorithm of reinforcement learning (RL), with the spectrum prediction error (RMSE) as the reward signal, the weight of the gating mechanism is updated every 1 minute, so that the prediction accuracy of the model in non-stationary sea conditions is increased by 18% compared with the static model.

[0167] In the embodiment of the present application, after receiving the real-time data uploaded by the edge node 701, the central node 702 can run a dynamic weight calculation engine. This engine mainly realizes the setting of basic weights, initializes weights for different service types (for example: emergency rescue weight 8, ocean monitoring weight 5, video monitoring weight 4, voice data weight 3), and realizes 1000 weight updates per second through FPGA acceleration.

[0168] The central node 702 can also adjust the weight based on historical BER and TD. When BER < 10 -5 and TD < 80ms, the weight is +2. When 10 -5 ≤ BER < 10 -4 and 80 ≤ TD < 150ms, the weight remains unchanged, otherwise the weight is -1. The distance d between the target device and the base station is obtained through Beidou differential positioning. When d ≤ 10 nautical miles, the weight remains unchanged. When 10 < d ≤ 20 nautical miles, the weight is +20%. When d > 20 nautical miles, the weight is +50%.

[0169] The central node 702 can also, when detecting an SOS signal, forcibly increase the weight of the emergency rescue service to 2 times that of the ordinary service and trigger a spectrum preemption process. Sort the communication requirements according to the comprehensive weight, preferentially allocate high-quality frequency bands (defined as frequency bands with SS-RSRP ≥ -98dBm and SS-SINR ≥ -3dB) for high-priority services, and monitor the frequency band load in real time. When the load > 80%, reduce the weight of the ships in the frequency band by 30% to release resources.

[0170] It should be noted that there seems to be an error in the original text where "every 10 minutes" in item [5] is translated as "every 1 minute" in the translation. It should be corrected according to the original meaning.The system provided in this application embodiment collects data from the target device in real time through edge nodes, extracts features from the data to obtain feature information, and generates an initial spectrum allocation scheme based on preset weights using a greedy algorithm according to the feature information. The control node obtains the initial scheme, divides the spectrum range in it according to a preset step size to obtain multiple candidate schemes, and determines the optimal scheme among the multiple candidate schemes based on the base station's spectrum remaining rate, communication quality indicators, and load balancing indicators. The system then allocates spectrum to the base station according to the optimal scheme, providing a dynamic spectrum allocation scheme that can improve spectrum utilization.

[0171] The aforementioned system employs a spatiotemporal joint neural network architecture with dynamic feature fusion. Through parallel branches of CNN and LSTM and adaptive gating mechanisms, it enhances the accuracy of spectral feature extraction and the model's generalization ability. By structurally innovating the traditional CNN-RNN fusion model, it achieves dynamic decoupling and fusion of spatiotemporal features of spectral data, multi-scale environmental awareness, and lightweight deployment, effectively addressing the problems of incomplete feature extraction and weak model generalization ability in complex sea conditions found in existing technologies. The dynamic feature fusion architecture, unlike traditional fixed-weight CNN-RNN combinations, introduces environmental condition input and an online self-optimization mechanism, endowing the model with the ability to adapt to changes in sea conditions. Based on boundary-preserving pooling, multi-scale attention, and low-rank regularization, it optimizes the feature extraction process for the characteristics of spectral data.

[0172] Figure 8 This is a schematic diagram of the hardware structure of an electronic device to implement the various embodiments of this application. The electronic device 800 includes, but is not limited to: a radio frequency unit 801, a network module 802, an audio output unit 803, an input unit 804, a sensor 805, a display unit 806, a user input unit 807, an interface unit 808, a memory 809, a processor 810, and a power supply 811, etc. Those skilled in the art will understand that... Figure 5 The electronic device structures shown are not intended to limit the electronic device. An electronic device may include more or fewer components than shown, or combine certain components, or have different component arrangements. In the embodiments of this application, the electronic device includes, but is not limited to, mobile phones, tablets, laptops, PDAs, in-vehicle terminals, wearable devices, and pedometers.

[0173] The processor 810 is used to obtain an initial scheme for spectrum allocation, the initial scheme being generated based on data collected in real time from the target device, the initial scheme including at least one spectrum range; the spectrum range is divided into multiple candidate schemes according to a preset step size; the optimal scheme is determined among the multiple candidate schemes based on the base station's spectrum remaining rate, communication quality indicators, and load balancing indicators; and spectrum is allocated to the base station according to the optimal scheme.

[0174] This application provides an electronic device that obtains an initial spectrum allocation scheme. This initial scheme is generated based on real-time collected data from the target device and includes at least one spectrum range. The spectrum range is divided into multiple candidate schemes according to a preset step size. Based on the base station's spectrum remaining rate, communication quality indicators, and load balancing indicators, the optimal scheme is determined among the multiple candidate schemes. The spectrum is allocated to the base station according to the optimal scheme, providing a dynamic spectrum allocation scheme that can improve spectrum utilization.

[0175] It should be understood that, in this embodiment, the radio frequency unit 801 can be used for receiving and transmitting signals during information transmission or calls. Specifically, it receives downlink data from the base station and processes it with the processor 810; additionally, it transmits uplink data to the base station. Typically, the radio frequency unit 801 includes, but is not limited to, an antenna, at least one amplifier, a transceiver, a coupler, a low-noise amplifier, a duplexer, etc. Furthermore, the radio frequency unit 801 can also communicate with networks and other electronic devices through a wireless communication system.

[0176] Electronic devices provide users with wireless broadband internet access through network module 802, such as helping users send and receive emails, browse web pages, and access streaming media.

[0177] The audio output unit 803 can convert audio data received by the radio frequency unit 801 or the network module 802 or stored in the memory 809 into audio signals and output them as sound. Furthermore, the audio output unit 803 can also provide audio output related to specific functions performed by the electronic device 800 (e.g., call signal reception sound, message reception sound, etc.). The audio output unit 803 includes a speaker, a buzzer, and a receiver, etc.

[0178] Input unit 804 is used to receive audio or video signals. Input unit 804 may include a graphics processing unit (GPU) 8041 and a microphone 8042. The GPU 8041 processes image data of still images or videos acquired by an image capture device (such as a camera) in video capture mode or image capture mode. The processed image frames can be displayed on display unit 806. The image frames processed by GPU 8041 can be stored in memory 809 (or other storage media) or transmitted via radio frequency unit 801 or network module 802. Microphone 8042 can receive sound and process such sound into audio data. The processed audio data can be converted into a format that can be transmitted to a mobile communication base station via radio frequency unit 801 in telephone call mode.

[0179] The electronic device 800 also includes at least one sensor 805, such as a light sensor, a motion sensor, and other sensors. Specifically, the light sensor includes an ambient light sensor and a proximity sensor. The ambient light sensor can adjust the brightness of the display panel 8061 according to the ambient light level, and the proximity sensor can turn off the display panel 8061 and / or backlight when the electronic device 800 is moved to the ear. As a type of motion sensor, an accelerometer sensor can detect the magnitude of acceleration in various directions (generally three axes). When stationary, it can detect the magnitude and direction of gravity and can be used to identify the posture of the electronic device (such as landscape / portrait switching, related games, magnetometer posture calibration), vibration recognition related functions (such as pedometer, tapping), etc. The sensor 805 may also include a fingerprint sensor, pressure sensor, iris sensor, molecular sensor, gyroscope, barometer, hygrometer, thermometer, infrared sensor, etc., which will not be described in detail here.

[0180] The display unit 806 is used to display information input by the user or information provided to the user. The display unit 806 may include a display panel 8061, which may be configured in the form of a liquid crystal display (LCD), an organic light-emitting diode (OLED), or the like.

[0181] User input unit 807 can be used to receive input numerical or character information, and to generate key signal inputs related to user settings and function control of electronic devices. Specifically, user input unit 807 includes a touch panel 8071 and other input devices 8072. Touch panel 8071, also known as a touch screen, can collect touch operations performed by the user on or near it (such as operations performed by the user using a finger, stylus, or any suitable object or accessory on or near touch panel 8071). Touch panel 8071 may include two parts: a touch detection device and a touch controller. The touch detection device detects the user's touch position and the signal generated by the touch operation, and transmits the signal to the touch controller; the touch controller receives touch information from the touch detection device, converts it into touch point coordinates, and sends it to processor 810, which receives and executes commands from processor 810. In addition, touch panel 8071 can be implemented using various types such as resistive, capacitive, infrared, and surface acoustic wave. Besides touch panel 8071, user input unit 807 may also include other input devices 8072. Specifically, other input devices 8072 may include, but are not limited to, physical keyboards, function keys (such as volume control buttons, power buttons, etc.), trackballs, mice, joysticks, etc., which will not be described in detail here.

[0182] Furthermore, the touch panel 8071 can cover the display panel 8061. When the touch panel 8071 detects a touch operation on or near it, it transmits the information to the processor 810 to determine the type of touch event. Subsequently, the processor 810 provides corresponding visual output on the display panel 8061 based on the type of touch event. Although in Figure 5 In this embodiment, the touch panel 8071 and the display panel 8061 are two independent components to realize the input and output functions of the electronic device. However, in some embodiments, the touch panel 8071 and the display panel 8061 can be integrated to realize the input and output functions of the electronic device. The specific implementation is not limited here.

[0183] Interface unit 808 serves as an interface for connecting external devices to electronic device 800. For example, external devices may include a wired or wireless headphone port, an external power supply (or battery charger) port, a wired or wireless data port, a memory card port, a port for connecting a device with an identification module, an audio input / output (I / O) port, a video I / O port, a headphone port, and so on. Interface unit 808 can be used to receive input from external devices (e.g., data, power, etc.) and transmit the received input to one or more components within electronic device 800, or it can be used to transmit data between electronic device 800 and external devices.

[0184] The memory 809 can be used to store software programs and various data. The memory 809 may primarily include a program storage area and a data storage area. The program storage area may store the operating system, applications required for at least one function (such as sound playback, image playback, etc.), etc.; the data storage area may store data created based on the use of the mobile phone (such as audio data, phonebook, etc.). Furthermore, the memory 809 may include high-speed random access memory, and may also include non-volatile memory, such as at least one disk storage device, flash memory device, or other volatile solid-state storage device.

[0185] The processor 810 is the control center of the electronic device. It connects various parts of the electronic device via various interfaces and lines. By running or executing software programs and / or modules stored in the memory 809, and by calling data stored in the memory 809, it performs various functions and processes data, thereby providing overall monitoring of the electronic device. The processor 810 may include one or more processing units; preferably, the processor 810 may integrate an application processor and a modem processor. The application processor mainly handles the operating system, user interface, and applications, while the modem processor mainly handles wireless communication. It is understood that the modem processor may not be integrated into the processor 810.

[0186] The electronic device 800 may also include a power supply 811 (such as a battery) that supplies power to various components. Preferably, the power supply 811 can be logically connected to the processor 810 through a power management system, thereby enabling functions such as managing charging, discharging, and power consumption through the power management system.

[0187] Preferably, this application embodiment also provides an electronic device, including a processor 810, a memory 809, and a computer program stored in the memory 809 and executable on the processor 810. When the computer program is executed by the processor 810, it implements the various processes of the above method embodiments and can achieve the same technical effect. To avoid repetition, it will not be described again here.

[0188] This application also provides a computer program product, which includes a computer program. When the computer program is executed by a processor, it implements the various processes of the above method embodiments and can achieve the same technical effect. To avoid repetition, it will not be described again here.

[0189] This application also provides a computer-readable storage medium storing a computer program. When executed by a processor, the computer program implements the various processes of the above-described method embodiments and achieves the same technical effects. To avoid repetition, it will not be described again here. The computer-readable storage medium may be a read-only memory (ROM), a random access memory (RAM), a magnetic disk, or an optical disk.

[0190] The computer-readable storage medium provided in this application embodiment obtains an initial spectrum allocation scheme. This initial scheme is generated based on data collected in real time from the target device and includes at least one spectrum range. The spectrum range is divided into multiple candidate schemes according to a preset step size. Based on the base station's spectrum remaining rate, communication quality indicators, and load balancing indicators, the optimal scheme is determined among the multiple candidate schemes. The spectrum is allocated to the base station according to the optimal scheme, thus providing a dynamic spectrum allocation scheme that can improve spectrum utilization.

[0191] Those skilled in the art will understand that embodiments of this application can be provided as methods, systems, or computer program products. Therefore, this application can take the form of a completely hardware embodiment, a completely software embodiment, or an embodiment combining software and hardware aspects. Furthermore, this application can take the form of a computer program product embodied on one or more computer-usable storage media (including but not limited to disk storage, CD-ROM, optical storage, etc.) containing computer-usable program code.

[0192] This application is described with reference to flowchart illustrations and / or block diagrams of methods, apparatus (systems), and computer program products according to embodiments of this application. It will be understood that each block of the flowchart illustrations and / or block diagrams, and combinations of blocks in the flowchart illustrations and / or block diagrams, can be implemented by computer program instructions. These computer program instructions can be provided to a processor of a general-purpose computer, special-purpose computer, embedded processor, or other programmable data processing apparatus to produce a machine, such that the instructions, which execute via the processor of the computer or other programmable data processing apparatus, generate instructions for implementing the flowchart... Figure 1 One or more processes and / or boxes Figure 1 A device that provides the functions specified in one or more boxes.

[0193] These computer program instructions may also be stored in a computer-readable storage medium that can direct a computer or other programmable data processing device to function in a particular manner, such that the instructions stored in the computer-readable storage medium produce an article of manufacture including instruction means, which are implemented in a process Figure 1 One or more processes and / or boxes Figure 1 The function specified in one or more boxes.

[0194] These computer program instructions may also be loaded onto a computer or other programmable data processing equipment to cause a series of operational steps to be performed on the computer or other programmable equipment to produce a computer-implemented process, thereby providing instructions that execute on the computer or other programmable equipment for implementing the process. Figure 1 One or more processes and / or boxes Figure 1 The steps of the function specified in one or more boxes.

[0195] In a typical configuration, a computing device includes one or more processors (CPU), input / output interfaces, network interfaces, and memory.

[0196] Memory may include non-persistent storage in computer-readable media, such as random access memory (RAM) and / or non-volatile memory, such as read-only memory (ROM) or flash RAM. Memory is an example of computer-readable media.

[0197] Computer-readable media includes both permanent and non-permanent, removable and non-removable media that can store information by any method or technology. Information can be computer-readable instructions, data structures, modules of programs, or other data. Examples of computer storage media include, but are not limited to, phase-change memory (PRAM), static random access memory (SRAM), dynamic random access memory (DRAM), other types of random access memory (RAM), read-only memory (ROM), electrically erasable programmable read-only memory (EEPROM), flash memory or other memory technologies, CD-ROM, digital versatile optical disc (DVD) or other optical storage, magnetic tape, magnetic magnetic disk storage or other magnetic storage devices, or any other non-transferable medium that can be used to store information accessible by a computing device. As defined herein, computer-readable media does not include transient computer-readable media, such as modulated data signals and carrier waves.

[0198] It should also be noted that the terms "comprising," "including," or any other variations thereof are intended to cover non-exclusive inclusion, such that a process, method, article, or apparatus that comprises a list of elements includes not only those elements but also other elements not expressly listed, or elements inherent to such a process, method, article, or apparatus. Without further limitation, an element defined by the phrase "comprising one..." does not exclude the presence of other identical elements in the process, method, article, or apparatus that includes said element.

[0199] The above description is merely an embodiment of this application and is not intended to limit the scope of this application. Various modifications and variations can be made to this application by those skilled in the art. Any modifications, equivalent substitutions, improvements, etc., made within the spirit and principles of this application should be included within the scope of the claims of this application.

Claims

1. A method for dynamically allocating spectrum, characterized in that, Applied to a control node, the method includes: An initial scheme for spectrum allocation is obtained, the initial scheme being generated based on data from the target device acquired in real time, and the initial scheme including at least one spectrum range; The spectrum range is divided into multiple candidate schemes according to a preset step size; Based on the base station's spectrum availability, communication quality indicators, and load balancing indicators, the optimal solution is determined from the multiple candidate solutions. Allocate spectrum to the base station according to the optimal scheme described above; The target device refers to one or more devices within the coverage area of ​​the base station.

2. The method according to claim 1, characterized in that, The process of determining the optimal solution from multiple candidate solutions based on base station spectrum availability, communication quality metrics, and load balancing metrics includes: The objective function of the candidate scheme is calculated based on the base station's spectrum availability, communication quality indicators, and load balancing indicators. The objective function difference is calculated based on the objective function of the candidate schemes. The acceptance probability is calculated based on the current temperature and the objective function difference. A random number is generated and compared with the acceptance probability to determine the current optimal solution of the objective function. The optimal solution is determined by iteratively analyzing the current temperature and cooling rate until the current temperature drops to the target temperature or the number of iterations reaches the target number. The global optimal solution of the objective function is then obtained as the optimal scheme for spectrum allocation.

3. The method according to claim 2, characterized in that, The objective function for calculating the candidate scheme based on the base station's spectrum surplus rate, communication quality indicators, and load balancing indicators includes: Based on the base station's spectrum availability, communication quality metrics, and load balancing metrics, the objective function of the candidate scheme is calculated using the following formula: ； ； ； ； ； Where f represents the objective function of the candidate scheme, R represents the spectrum surplus rate, Q represents the communication quality index, and B represents the load balancing index. Indicates the total bandwidth of the available spectrum. Indicates the allocated spectrum bandwidth. This represents the average bit error rate. The expression represents the average transmission delay, 'a' represents the bit error rate threshold, 'b' represents the transmission delay threshold, and 'n' represents the number of base stations. This indicates the load of each base station. This indicates the average load of the base station. The weights representing the spectrum surplus rate The weights of communication quality metrics are indicated. This indicates the weight of the load balancing metric.

4. The method according to claim 2, characterized in that, The step of calculating the objective function difference based on the objective function of the candidate schemes, calculating the acceptance probability based on the current temperature and the objective function difference, generating a random number and comparing it with the acceptance probability to determine the current optimal solution of the objective function includes: Determine the current candidate solution and alternative candidate solutions from the plurality of candidate solutions; Calculate the difference in objective functions based on the objective functions of the current candidate solution and the alternative candidate solutions; The acceptance probability P is calculated based on the difference between the current temperature and the objective function. Generate a random number r. If r ≤ P, then the candidate solution is taken as the current optimal solution of the objective function. If r > P, then the current candidate solution is taken as the current optimal solution of the objective function. Where 0 ≤ r ≤ 1.

5. The method according to claim 1, characterized in that, Also includes: In the event of an emergency, the optimal solution shall be adjusted according to at least one of the following: Temporarily allocate idle frequency bands and / or reserve frequency bands for the aforementioned emergency services; Increase the weight of the aforementioned emergency services; The communication quality requirements for non-emergency services will be temporarily relaxed.

6. The method according to claim 1, characterized in that, It also includes at least one of the following: If the load on a certain frequency band in the optimal solution exceeds the load threshold, the weight of ordinary services in that frequency band is reduced, and idle frequency bands or frequency bands occupied by low priority are reclaimed. Adjustments will be made to the reserved global frequency bands during peak communication periods or important business operations. When there are multiple base stations and their coverage areas overlap, the spectrum in the optimal scheme is set to be shared by the multiple base stations.

7. The method according to claim 1, characterized in that, Also includes: Real-time monitoring of synchronization signal reference signal received power (SS-RSRP), synchronization signal signal-to-interference-plus-noise ratio (SS-SINR), and bit error rate; If the SS-RSRP is lower than the first threshold, then the dynamic power adjustment algorithm is used to optimize the transmitter parameters; If the SS-SINR is lower than the second threshold, then adaptive filtering techniques are used to suppress noise; If the bit error rate is higher than the third threshold, forward error correction coding is used for correction or retransmission is performed.

8. The method according to claim 1, characterized in that, Also includes: Calculate the spectrum hole rate based on the bandwidth idle ratio threshold; Calculate the spectrum switching frequency of the base station; If the spectrum switching frequency exceeds the switching threshold, a switching suppression algorithm is used to reduce the spectrum switching frequency. When the load of the base station exceeds the load threshold, the bandwidth idle ratio threshold is reduced to adjust the spectrum hole rate, and the upper limit threshold of the spectrum switching frequency is increased.

9. The method according to claim 8, characterized in that, The calculation of the spectrum hole rate based on the bandwidth idle ratio threshold includes: The spectral void ratio within the statistical period is calculated using the following formula: ； ； ； in, This represents the duration of time the k-th frequency band is occupied within the statistical period. This represents the end time of the i-th use of the k-th frequency band within the statistical period. This represents the start time of the i-th use of the k-th frequency band within the statistical period, and N represents the total number of times the k-th frequency band is used within the statistical period. This represents the idle time of the k-th frequency band within the statistical period. Indicates the statistical period. Indicates the spectral void ratio. Indicates the threshold for continuous idle time. This indicates the threshold for the proportion of bandwidth that is not in use. This indicates an exponential function; the exponential function takes the value 1 if the condition within the parentheses is met, and takes the value 0 if the condition within the parentheses is not met. This represents the total bandwidth of the k-th frequency band within the statistical period. This represents the idle bandwidth of the k-th frequency band within the statistical period, where K represents the number of frequency bands.

10. The method according to claim 1, characterized in that, Also includes: The activity level of each frequency band within the statistical period is calculated using the following formula: ； in, This indicates the activity level of the k-th frequency band within the statistical period. This indicates the number of times the k-th frequency band is used within the statistical period. This represents the bandwidth of the k-th frequency band within the statistical period. Indicates the statistical period; If the activity level is greater than the activity threshold, the k-th frequency band is split into multiple sub-frequency bands for allocation.

11. The method according to claim 1, characterized in that, Also includes: Real-time monitoring of the speed of the target device; If the target device's moving speed exceeds a speed threshold, reduce the spectrum sensing period and reduce the continuous idle time threshold. The Used to calculate the spectral hole rate.

12. A device for dynamically allocating spectrum, characterized in that, Applied to a control node, the device includes: An acquisition module is used to acquire an initial scheme for spectrum allocation, the initial scheme being generated based on data collected in real time from the target device, and the initial scheme including at least one spectrum range; The partitioning module is used to partition the spectrum range according to a preset step size to obtain multiple candidate schemes; The determination module is used to determine the optimal solution among the multiple candidate solutions based on the base station's spectrum remaining rate, communication quality indicators, and load balancing indicators. An allocation module is used to allocate spectrum to the base station according to the optimal scheme; The target device refers to one or more devices within the coverage area of ​​the base station.

13. A system for dynamically allocating spectrum, characterized in that, The system includes: Edge nodes are used to collect data from target devices in real time, and to extract features from the data to obtain feature information, which includes at least one of frequency domain features, time features, and spatial features. A central node is used to generate an initial spectrum allocation scheme based on preset weights using a greedy algorithm according to the feature information obtained from the edge nodes. The initial scheme includes at least one spectrum range. A control node is used to obtain the initial scheme generated by the central node, divide the spectrum range into multiple candidate schemes according to a preset step size, determine the optimal scheme among the multiple candidate schemes based on the base station's spectrum remaining rate, communication quality indicators and load balancing indicators, and allocate spectrum to the base station according to the optimal scheme. The target device refers to one or more devices within the coverage area of ​​the base station.

14. The system according to claim 13, characterized in that, The edge nodes extract feature information from the data, including: If the proportion of missing data in the data is less than or equal to a preset proportion, the missing data is filled with the mean or median of the non-missing data. If the proportion of missing data in the data is greater than the preset proportion, the missing data is predicted and filled using a time series model or a machine learning regression model. Remove noisy data from the filled data; Feature information is obtained by extracting features from the data after noise removal. Each parameter in the feature information is labeled based on its corresponding initial threshold, and the label is marked as excellent or poor.

15. The system according to claim 13, characterized in that, The edge nodes use a neural network model to extract features from the data to obtain feature information, wherein the neural network model includes: CNN layers are used to extract features from input data and output frequency domain features; Pooling layers are used to transform the frequency domain features; LSTM layers are used to extract features from input data and output temporal and spatial features. The first fully connected layer is used to perform multi-dimensional semantic parsing of the frequency domain features, time domain features and spatial features according to multiple preset semantic subspaces, and learn feature associations; The second fully connected layer is used to compress the feature dimension of the output of the first fully connected layer.

16. An electronic device, characterized in that, It includes a processor, a memory, and a computer program stored in the memory and executable on the processor, wherein the computer program, when executed by the processor, implements the steps of the method for dynamically allocating spectrum as described in any one of claims 1-11.

17. A computer-readable storage medium, characterized in that, A computer program is stored on the computer-readable storage medium, which, when executed by a processor, implements the steps of the method for dynamically allocating spectrum as described in any one of claims 1-11.

18. A computer program product, characterized in that, Includes a computer program that, when executed by a processor, implements the steps of the method for dynamically allocating spectrum as described in any one of claims 1-11.

Images