A broadband spectrum monitoring system based on dynamic channel allocation

Abstract

The present invention relates to the field of broadband spectrum technology and discloses a broadband spectrum monitoring system based on dynamic channel allocation, comprising a data acquisition unit, a data identification unit, a region division unit, a data alignment unit, and a channel allocation unit. The system dynamically adjusts the coverage range through beacon nodes, accurately calculates the coverage radius in combination with the urban environment, and optimizes the channel allocation strategy using game theory. The system responds to changes in the spectrum environment and integrates multi-dimensional features to construct spectrum state parameters, ensuring smooth communication, eliminating blind spots of monitoring nodes, and achieving data spatiotemporal alignment, effectively improving monitoring sensitivity and resource utilization, ensuring spectrum stability and reliable communication during large-scale activities in the city, and guaranteeing dynamic and reasonable channel allocation.

Description

Technical Field

[0001] The present invention relates to the technical field of broadband spectrum, and in particular to a broadband spectrum monitoring system based on dynamic channel allocation. Background Art

[0002] When existing broadband spectrum monitoring systems are used in cities, they face the spectrum challenges brought about by large-scale urban events. They mainly adopt a combination of historical data-based predictions and distributed spectrum monitoring to achieve monitoring and channel allocation. Usually, they first collect spectrum usage data of historical activities in the area, and use machine learning algorithms to predict spectrum demand in different time periods and different areas during the event. At the same time, through a distributed spectrum monitoring network, multiple monitoring nodes are deployed in the city. These nodes can scan the spectrum in their area in real time, collecting data such as spectrum occupancy and signal strength to achieve the purpose of activity monitoring.

[0003] However, in complex large-scale urban activity scenarios, blind spots are prone to occur in the range of distributed monitoring nodes, and there is a lack of effective coordination mechanisms when fusing data from each node, resulting in data inconsistency. This reduces the monitoring sensitivity to drastic changes in spectrum usage in local areas, making it impossible to detect spectrum anomalies in a timely and accurate manner, leading to unreasonable channel allocation. Summary of the Invention

[0004] In view of the deficiencies of the prior art, the present invention provides a broadband spectrum monitoring system based on dynamic channel allocation to solve the above problems.

[0005] The above technical objectives of the present invention are achieved through the following technical solutions:

[0006] A broadband spectrum monitoring system based on dynamic channel allocation, comprising:

[0007] A data acquisition unit is used to establish multiple monitoring nodes in the city, divide the city into multiple monitoring areas, and divide the monitoring nodes into beacon nodes and ordinary nodes. The beacon nodes obtain spectrum monitoring data and historical spectrum monitoring data of the monitoring area in real time, dynamically adjust the beacon node's transmission power based on the real-time spectrum monitoring data, synchronously calculate the effective coverage radius parameter of the beacon node, and calculate the signal strength and occupancy time data of each frequency band in the current monitoring area collected in real time by the ordinary nodes to obtain a set of frequency band characteristic parameters;

[0008] The data identification unit is used to adjust the monitoring range of ordinary nodes according to the effective coverage radius parameter, extract the frequency band characteristic parameter set, and generate a shared parameter set including the location and occupancy of the spectrum hole;

[0009] The area division unit is used to divide the monitoring area into multiple game areas, take the beacon node in each area as the game subject, calculate the channel allocation benefit parameters of the shared parameter set, and obtain the final channel allocation strategy parameters of each game area;

[0010] A data alignment unit is used to perform spatiotemporal alignment processing on the real-time spectrum monitoring data of the monitoring nodes in each game area according to the final channel allocation strategy parameters, and generate global spectrum state parameters including signal strength gradient distribution and spectrum state mutation probability;

[0011] The channel allocation unit is used to determine whether the current channel allocation strategy is invalid based on the global spectrum status parameters, and immediately trigger the dynamic allocation instruction if invalid.

[0012] Furthermore, the transmit power of the beacon node is dynamically adjusted based on the real-time spectrum monitoring data, and the effective coverage radius parameters of the beacon node are synchronously calculated, including:

[0013] For each frequency band in the real-time spectrum monitoring data, calculate its and The absolute value of the signal strength difference over time is divided by the dynamic noise floor of the frequency band, and then the sum is calculated for all frequency bands to obtain the total parameter of the change of the signal strength of each frequency band relative to the noise floor;

[0014] For each frequency band, first calculate the time derivative of the signal strength and multiply it by Then divide it by the mean signal strength of the entire frequency band to obtain the time derivative parameter of the signal strength after frequency band normalization;

[0015] After squaring the time derivative parameter of the signal strength after frequency band normalization, the sum of all frequency bands is added to 1 to obtain the comprehensive parameter of the relationship between the square of the signal strength change rate and the mean value of all frequency bands;

[0016] The sum parameter is divided by the comprehensive parameter, and then the result is calculated by the hyperbolic tangent function to obtain the dynamic interference compensation coefficient;

[0017] The dynamic interference compensation coefficient is combined with the basic transmit power of the beacon node to obtain the current actual transmit power value;

[0018] The current actual transmission power value, urban building density and wireless propagation loss are integrated and calculated to generate the effective coverage radius parameter.

[0019] Furthermore, the signal strength and occupancy time data of each frequency band in the current monitoring area collected in real time by ordinary nodes are calculated to obtain a set of frequency band characteristic parameters, including:

[0020] Cluster the signal strength data collected by adjacent common nodes in the same time period to generate a spatiotemporal dynamic feature matrix;

[0021] By analyzing the spatiotemporal dynamic characteristic matrix, the signal strength fluctuation variance, multipath fading factor and Doppler frequency shift rate of the wireless channel are obtained;

[0022] Calculate the ratio of the standard deviation to the mean of the signal strength in the frequency band, calculate the ratio of the Doppler frequency shift rate to the maximum Doppler frequency shift rate supported by the system, and multiply the two ratios to obtain the fluctuation-frequency shift coupling factor;

[0023] Take the square root of the multipath fading factor to get the multipath fading adjustment factor;

[0024] Divide the fluctuation-frequency shift coupling factor by the multipath fading adjustment factor to obtain the signal dynamic characteristic index;

[0025] Then calculate the ratio of the maximum and minimum signal strengths of the frequency band in the monitoring area to obtain the signal strength extreme value ratio, and multiply the signal strength extreme value ratio by the signal dynamic characteristic index to obtain the signal quality factor;

[0026] The traditional occupancy time of a specific frequency band monitored by ordinary nodes is weighted and calculated according to the signal quality factor to generate an adaptive occupancy rate;

[0027] Calculate the spatiotemporal dynamic feature matrix, adaptive occupancy rate and dynamic interference compensation coefficient of beacon nodes to generate a three-dimensional feature vector;

[0028] Perform dimensionality reduction on the principal components of the three-dimensional feature vector to obtain a triplet of feature parameters including frequency band activity, stability, and interference sensitivity;

[0029] The feature screening threshold is dynamically adjusted according to the current network load, the local density and distance of the feature parameters in the feature parameter triplet are calculated, and a frequency band feature parameter set is generated.

[0030] Furthermore, the monitoring range of common nodes is adjusted according to the effective coverage radius parameter, including:

[0031] Calculate the monitoring range adjustment coefficient of ordinary nodes based on the effective coverage radius parameters of the beacon node and the monitoring area overlap requirements;

[0032] The original monitoring range of the ordinary node is combined with the monitoring range adjustment coefficient to generate a new monitoring range, specifically including: multiplying the initial monitoring range of the ordinary node by the monitoring range adjustment coefficient to obtain the adjusted monitoring radius; through the three-dimensional space buffer analysis algorithm, with the coordinates of the ordinary node as the center and the adjusted monitoring radius as the new radius, a spherical basic monitoring area is generated, which is the new monitoring range.

[0033] Furthermore, the frequency band characteristic parameter set is extracted to generate a shared parameter set containing the spectrum hole location and occupancy, including:

[0034] According to the new monitoring range, the signal strength and occupancy time data of each frequency band in the dynamic monitoring area are extracted from the frequency band characteristic parameter set to generate spatiotemporal characteristic correlation parameters;

[0035] Based on the spatiotemporal feature correlation parameters, the signal strength threshold and the occupancy time threshold are set, and the areas with signal strength lower than the signal strength threshold and occupancy time lower than the occupancy time threshold are screened out to obtain the spectrum hole position parameters;

[0036] For the monitoring area corresponding to the spectrum hole location parameter, the proportion of the occupation time of each frequency band in the monitoring area to the total monitoring time is calculated to obtain the spectrum hole occupancy rate parameter. Specifically, the method includes: using a spatial overlay analysis algorithm to spatially match the node geographic coordinates with the continuous area of the spectrum hole, extracting the node ID and corresponding frequency band falling within the area, grouping by frequency band, summarizing the weighted occupation time of each frequency band in all time windows, and calculating the total occupation time of the frequency band in the monitoring area. The total monitoring time is the total time of the same period of the spatiotemporal dynamic feature matrix. Using a data aggregation algorithm, the total occupation time of each frequency band is divided by the total monitoring time to obtain the spectrum hole occupancy rate parameter.

[0037] The spectrum hole location parameters and the spectrum hole occupancy parameters are integrated to generate a shared parameter set including the spectrum hole location and occupancy.

[0038] Furthermore, the channel allocation benefit parameters of the shared parameter set are calculated to obtain the final channel allocation strategy parameters of each game area, including:

[0039] Based on the spectrum hole locations and occupancy rates in the shared parameter set, the availability of each frequency band is determined and a frequency band availability parameter is generated.

[0040] The frequency band availability parameters are combined with the spectrum requirements of each gaming area and analyzed to generate a preliminary channel allocation strategy;

[0041] Calculate the benefits of the preliminary channel allocation strategy to obtain the channel allocation benefit parameters;

[0042] The channel allocation benefit parameters are calculated to obtain the final channel allocation strategy parameters of each game area.

[0043] Furthermore, the real-time spectrum monitoring data of the monitoring nodes in each game area are temporally and spatially aligned according to the final channel allocation strategy parameters to generate global spectrum state parameters including signal strength gradient distribution and spectrum state mutation probability, including:

[0044] Based on the time slice division in the final channel allocation strategy parameters, an adaptive sliding time window is generated for each monitoring node. The signal propagation delay is calculated and the time window boundary is dynamically adjusted through the time-standard signal transmitted by the beacon node and the received signal strength of the ordinary node to generate the time-space synchronization calibration parameters.

[0045] Dynamically adjust the weight index of common monitoring nodes according to the monitoring node distribution density and beacon node effective coverage radius parameters to generate signal strength gradient distribution parameters;

[0046] The spatiotemporal synchronization calibration parameters and the spectrum state differences of adjacent time slices are analyzed to obtain the probability of spectrum state mutation;

[0047] The signal intensity gradient distribution parameters and the spectrum state mutation probability are analyzed to generate global spectrum state parameters including the signal intensity gradient distribution and the spectrum state mutation probability.

[0048] Furthermore, based on the global spectrum status parameters, it is determined whether the current channel allocation strategy is invalid. If invalid, a dynamic allocation instruction is immediately triggered, including:

[0049] The historical spectrum monitoring data is divided into k time slices and frequency bands. The gradient mean of the spatial distribution of the signal strength of each frequency band and the rate of change of the spectrum occupancy in the time dimension are calculated slice by slice, forming a dynamic feature sequence that contains the spatiotemporal variation law.

[0050] The data of the time period corresponding to the historical channel allocation is grouped by the game area, and the median of the signal strength gradient and the extreme value of the spectrum occupancy change rate in the dynamic feature sequence of each monitoring area are counted respectively to generate preliminary regional gradient benchmark parameters and regional mutation benchmark parameters;

[0051] Based on the building density of the current monitoring area, the preliminary regional gradient benchmark parameters and regional mutation benchmark parameters are modified to obtain the final signal intensity gradient threshold and mutation probability threshold that are adapted to the current monitoring environment.

[0052] Furthermore, judging whether the current channel allocation strategy is invalid according to the global spectrum status parameters, and immediately triggering the dynamic allocation instruction if invalid, further comprising:

[0053] If the signal strength gradient in the global spectrum state parameter is greater than the signal strength gradient threshold or the mutation probability is continuously higher than the mutation probability threshold for k consecutive time slices, the previous channel allocation strategy is determined to be invalid and the dynamic allocation instruction is triggered immediately.

[0054] Furthermore, dynamic allocation instructions include:

[0055] The area where the signal intensity gradient exceeds the signal intensity gradient threshold is marked as an abnormal area, and the spatiotemporal correlation parameters between the abnormal area and the adjacent areas are calculated to generate the regional abnormal propagation factor;

[0056] Analyze the signal strength gradient distribution of each monitoring node in the abnormal area to obtain the spectrum state chaos degree;

[0057] According to the regional abnormal propagation factor and the disorder degree of the spectrum state, the boundaries of the game area are dynamically adjusted to generate a temporary spectrum management area;

[0058] Analyze the monitoring nodes within the temporary spectrum management area to obtain the channel resource desirability vector;

[0059] Extract regional abnormal propagation factors, spectrum state chaos, and channel resource demand vectors to obtain multi-dimensional risk assessment parameters;

[0060] When the evaluation value of any dimension of the multidimensional risk evaluation parameters exceeds the multidimensional risk threshold, a dynamic allocation instruction including the optimal channel combination, power adjustment and time slice allocation strategy is generated.

[0061] In summary, the present invention mainly has the following beneficial effects:

[0062] By establishing multiple monitoring nodes in a city, including beacon nodes and ordinary nodes, and adjusting the monitoring range of ordinary nodes based on the effective coverage radius parameters of the beacon nodes, the problem of blind spots easily generated by distributed monitoring nodes can be effectively solved. In the existing technology, distributed monitoring nodes operate independently, which easily leads to areas that cannot be monitored. In this system, beacon nodes obtain spectrum monitoring data and historical spectrum monitoring data in real time, dynamically adjust their own transmission power, and then calculate the effective coverage radius based on the density of urban buildings and wireless propagation loss. This provides a basis for accurately adjusting the monitoring range of ordinary nodes, enabling the entire monitoring network to work closely together, expand the effective monitoring area, and reduce blind spots. At the same time, after adjusting the monitoring range of ordinary nodes based on the effective coverage radius parameters, the data identification unit extracts the frequency band characteristic parameter set and generates a shared parameter set containing the location and occupancy of spectrum holes. This allows the data collected by each node to be processed and integrated within a unified monitoring range framework, avoiding the data inconsistency caused by inconsistent node monitoring ranges during previous data fusion. This improves the accuracy and reliability of the data, thereby enhancing the monitoring sensitivity to rapid changes in spectrum usage in local areas and enabling timely and accurate detection of spectrum anomalies.

[0063] By dividing the monitoring area into multiple game areas, with the beacon nodes in the area as the game subjects, the channel allocation benefit parameters of the shared parameter set are calculated to obtain the final channel allocation strategy parameters of each game area. This fully considers the actual spectrum usage and potential benefits of different areas, avoiding the unreasonable channel allocation problem caused by the lack of an effective coordination mechanism in traditional systems. The data alignment unit performs spatiotemporal alignment processing on the real-time spectrum monitoring data of the monitoring nodes in each game area based on the final channel allocation strategy parameters, and generates global spectrum state parameters including signal strength gradient distribution and spectrum state mutation probability, providing comprehensive and accurate spectrum state information for the channel allocation unit. The channel allocation unit uses this information to determine whether the current channel allocation strategy is invalid. If invalid, it immediately triggers dynamic allocation instructions to achieve dynamic optimization allocation of channel resources, ensure that spectrum resources can be flexibly adjusted according to actual usage and needs, improve spectrum resource utilization, meet the ever-changing spectrum needs in complex scenarios such as large-scale urban events, and ensure smooth and efficient communication.

[0064] By considering multiple factors in real-time spectrum monitoring data, including signal strength variations, the dynamic noise floor, and the time derivative of signal strength, the system accurately reflects the interference situation in the current spectrum environment. Based on this information, it dynamically adjusts the transmit power of beacon nodes and calculates the effective coverage radius, enabling the entire monitoring system to promptly adapt to changes in the spectrum environment. The generation of the frequency band characteristic parameter set incorporates multi-dimensional information, including the spatiotemporal dynamic feature matrix, adaptive occupancy, and the dynamic interference compensation coefficient of the beacon nodes. Through dimensionality reduction and feature filtering, a triplet of characteristic parameters, including band activity, stability, and interference sensitivity, is derived. This constructs a comprehensive and accurate description of the frequency band characteristics, providing strong support for subsequent channel allocation strategy optimization. When global spectrum state parameters exhibit anomalies, such as a signal strength gradient exceeding a threshold or a mutation probability consistently exceeding a threshold, the system quickly determines that the channel allocation strategy has failed and triggers dynamic allocation instructions. This rapid response mechanism enhances the system's adaptability and flexibility in the face of sudden spectrum changes and complex dynamic spectrum environments, ensuring that the system can quickly and accurately respond to spectrum challenges encountered in large-scale urban events and maintain stable and reliable spectrum usage. BRIEF DESCRIPTION OF THE DRAWINGS

[0065] Figure 1 It is a block diagram of the broadband spectrum monitoring system based on dynamic channel allocation of the present invention. DETAILED DESCRIPTION

[0066] The following will clearly and completely describe the technical solutions in the embodiments of the present invention in conjunction with the accompanying drawings. Obviously, the described embodiments are only part of the embodiments of the present invention, not all of the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by ordinary technicians in this field without making creative efforts are within the scope of protection of the present invention.

[0067] refer to Figure 1 , a broadband spectrum monitoring system based on dynamic channel allocation, comprising:

[0068] A data acquisition unit is used to establish multiple monitoring nodes in the city, divide the city into multiple monitoring areas, and divide the monitoring nodes into beacon nodes and ordinary nodes. The beacon nodes obtain spectrum monitoring data and historical spectrum monitoring data of the monitoring area in real time, dynamically adjust the beacon node's transmission power based on the real-time spectrum monitoring data, synchronously calculate the effective coverage radius parameter of the beacon node, and calculate the signal strength and occupancy time data of each frequency band in the current monitoring area collected in real time by the ordinary nodes to obtain a set of frequency band characteristic parameters;

[0069] The data identification unit is used to adjust the monitoring range of ordinary nodes according to the effective coverage radius parameter, extract the frequency band characteristic parameter set, and generate a shared parameter set including the location and occupancy of the spectrum hole;

[0070] The area division unit is used to divide the monitoring area into multiple game areas, take the beacon node in each area as the game subject, calculate the channel allocation benefit parameters of the shared parameter set, and obtain the final channel allocation strategy parameters of each game area;

[0071] A data alignment unit is used to perform spatiotemporal alignment processing on the real-time spectrum monitoring data of the monitoring nodes in each game area according to the final channel allocation strategy parameters, and generate global spectrum state parameters including signal strength gradient distribution and spectrum state mutation probability;

[0072] The channel allocation unit is used to determine whether the current channel allocation strategy is invalid based on the global spectrum status parameters, and immediately trigger the dynamic allocation instruction if invalid.

[0073] By setting up multiple monitoring nodes and dividing them into beacon nodes and ordinary nodes, spectrum monitoring data of various areas in the city can be accurately obtained. The beacon node dynamically adjusts the transmission power based on the data, determines the effective coverage radius, and provides a data basis for subsequent monitoring. Secondly, the data identification unit adjusts the monitoring range and extracts the frequency band characteristics to generate a shared parameter set to help identify spectrum holes and analyze occupancy. The regional division unit introduces game theory and calculates the final channel allocation strategy based on the beacon node to improve the rationality of spectrum resource allocation. The data alignment unit performs time-space alignment processing to generate global spectrum state parameters and enhance the accuracy of spectrum state perception. The channel allocation unit triggers dynamic allocation in a timely manner based on global parameters to ensure efficient and stable operation of the communication system, effectively improve spectrum utilization, provide strong support for broadband spectrum monitoring and channel allocation, and ensure smooth urban communications.

[0074] In one scenario of this embodiment, dynamically adjusting the transmit power of the beacon node according to real-time spectrum monitoring data and synchronously calculating the effective coverage radius parameter of the beacon node include:

[0075] For each frequency band in the real-time spectrum monitoring data, calculate its and The absolute value of the signal strength difference over time is divided by the dynamic noise floor of the frequency band, and then the sum is calculated for all frequency bands to obtain the total parameter of the change of the signal strength of each frequency band relative to the noise floor;

[0076] For each frequency band, first calculate the time derivative of the signal strength and multiply it by Then divide it by the mean signal strength of the entire frequency band to obtain the time derivative parameter of the signal strength after frequency band normalization;

[0077] After squaring the time derivative parameter of the signal strength after frequency band normalization, the sum of all frequency bands is calculated. 1. Obtain the comprehensive parameter of the relationship between the square of the signal strength change rate and the mean value of the entire frequency band;

[0078] The sum parameter is divided by the comprehensive parameter, and then the result is calculated by the hyperbolic tangent function to obtain the dynamic interference compensation coefficient;

[0079] When applied in practice, the specific calculation formula of the above dynamic interference compensation coefficient is as follows:

[0080] ;

[0081] Where, represents the dynamic interference compensation coefficient, represents the hyperbolic tangent function, Indicates frequency band In time The instantaneous signal strength, Indicates frequency band The dynamic noise floor, represents the length of the dynamic time window, , represents the time derivative of signal strength, Indicates the average signal strength of the entire frequency band;

[0082] The dynamic interference compensation coefficient is integrated with the basic transmission power of the beacon node to obtain the current actual transmission power value, specifically including: normalizing the dynamic interference compensation coefficient, and then multiplying the dynamic interference compensation coefficient by the dynamic interference compensation coefficient to obtain the current actual transmission power value;

[0083] The current actual transmission power value, urban building density and wireless propagation loss are integrated and calculated to generate the effective coverage radius parameter, which specifically includes: based on the current actual transmission power value, spatial interpolation processing is performed on the urban building density data to generate a continuously distributed three-dimensional density cloud map, and the three-dimensional density cloud map is converted into a signal propagation obstacle grid map through a rasterization projection algorithm. Each grid is assigned a corresponding attenuation weight value, and a Monte Carlo simulation algorithm is used to randomly generate a large number of signal propagation paths in the grid map. The cumulative attenuation value of each path is counted, and based on the sensitivity of the receiving device, the paths whose attenuation value does not exceed the difference between the transmission power and the receiving sensitivity are screened out, and the statistical average of the maximum propagation distances of these paths is taken as the effective coverage radius parameter.

[0084] The transmission power of the beacon node is dynamically adjusted based on real-time spectrum monitoring data, and the effective coverage radius parameters are calculated synchronously. The dynamic interference compensation coefficient is generated by extracting real-time spectrum data and integrated with the basic transmission power, so that the transmission power of the beacon node can adapt to changes in the spectrum environment in real time, improve the monitoring sensitivity to drastic changes in spectrum usage in local areas, and can detect spectrum anomalies in a timely and accurate manner, thereby providing a strong basis for the reasonable allocation of channels, thereby improving the utilization efficiency of spectrum resources and ensuring the stability and reliability of wireless communications in urban activity scenarios.

[0085] By calculating the actual transmission power by fusing the dynamic interference compensation coefficient with the basic transmission power, and integrating the urban building density and wireless propagation loss to generate the effective coverage radius parameter, the communication performance of beacon nodes in complex urban environments can be effectively improved. On the one hand, by accurately compensating for dynamic interference, the transmit power adaptability of the beacon node is improved, enabling it to maintain good communication quality in a changing spectrum environment. On the other hand, by comprehensively considering the urban building density and wireless propagation loss, the generated effective coverage radius parameter is more in line with the actual propagation environment, providing an accurate reference for the layout and optimization of beacon nodes, and thus rationally planning the wireless communication network, avoiding resource waste and interference problems caused by insufficient or excessive coverage, and further improving the overall performance and stability of the wireless communication network in urban activity scenarios.

[0086] In one case of this embodiment, the signal strength and occupancy duration data of each frequency band in the current monitoring area collected in real time by the common node are calculated to obtain a set of frequency band characteristic parameters, including:

[0087] The signal strength data collected by adjacent common nodes in the same time period are clustered to generate a spatiotemporal dynamic feature matrix. Specifically, the method includes: determining the set of adjacent common nodes based on geographic coordinates, aligning the signal strength data in the same time period by timestamps, segmenting the data according to fixed time windows, and using a spatiotemporal clustering algorithm to identify spatiotemporal clusters with similar signal strengths for the signal strength sequences of adjacent nodes in each window. The signal strength mean, variance, number of nodes covered by the cluster, and proportion of cluster duration within the spatiotemporal cluster are extracted, and the clusters are arranged in chronological order. The spatiotemporal dynamic feature matrix is generated using node ID, frequency band, and timestamp as indexes.

[0088] The spatiotemporal dynamic feature matrix is analyzed to obtain the signal strength fluctuation variance, multipath fading factor and Doppler frequency shift rate of the wireless channel. Specifically, the spatiotemporal dynamic feature matrix is grouped in two dimensions according to node ID and frequency band, and the signal strength mean sequence of each node in different time windows under each frequency band is extracted. The variance of the sequence on the time axis is calculated and used as the signal strength fluctuation variance. Based on the statistical distribution characteristics of the signal strength in the spatiotemporal dynamic feature matrix, the maximum likelihood estimation algorithm is used to fit the distribution parameters. The direct path power (obtained by the square of the signal strength mean combined with the noise power calibration) is calculated and the scattered path power is obtained. Path power (the power ratio of the two is obtained by subtracting the noise power from the signal strength variance). The scattered path power can be used as the multipath fading factor. Fast Fourier transform is performed on the signal data of adjacent time windows to convert the time domain signal into a frequency domain spectrum. The main peak frequency position in the spectrum is detected using spectral analysis. The offset of the main peak frequency of adjacent windows is calculated and combined with the real-time geographic coordinates of the node to calculate the relative motion speed of the node and the signal source. The frequency offset is associated with the relative motion speed through the frequency shift estimation algorithm. The Doppler frequency shift rate is directly obtained through the preset signal carrier frequency parameter (that is, the fixed signal operating frequency of the transmitter).

[0089] Calculate the ratio of the standard deviation to the mean of the signal strength in the frequency band, calculate the ratio of the Doppler frequency shift rate to the maximum Doppler frequency shift rate supported by the system, and multiply the two ratios to obtain the fluctuation-frequency shift coupling factor;

[0090] Take the square root of the multipath fading factor to get the multipath fading adjustment factor;

[0091] Divide the fluctuation-frequency shift coupling factor by the multipath fading adjustment factor to obtain the signal dynamic characteristic index;

[0092] Then calculate the ratio of the maximum and minimum signal strengths of the frequency band in the monitoring area to obtain the signal strength extreme value ratio, and multiply the signal strength extreme value ratio by the signal dynamic characteristic index to obtain the signal quality factor;

[0093] When applied in a specific application, the specific calculation formula of the above signal quality factor is as follows:

[0094] ;

[0095] Where, represents the signal quality factor, Indicates frequency band The standard deviation of the signal strength, Indicates frequency band The mean signal strength, represents the Doppler shift rate, Indicates the maximum Doppler frequency shift rate supported by the system. represents the multipath fading factor, Indicates the ratio of extreme signal strength values within the monitoring area;

[0096] The traditional occupancy time of a specific frequency band monitored by ordinary nodes is weighted and calculated according to the signal quality factor to generate an adaptive occupancy rate, specifically including: based on the signal quality factor, the specific frequency band is The traditional occupancy time is divided into time windows (with the same period as the spatiotemporal dynamic feature matrix), and the signal quality factor value in each window is extracted as the weighting coefficient. The occupancy time of each window is multiplied by the corresponding signal quality factor value and then accumulated to obtain the weighted total occupancy time. After normalization, the ratio of the weighted total time to the total monitoring time of the frequency band (the sum of all window times) is taken as the adaptive occupancy rate;

[0097] The spatiotemporal dynamic feature matrix, adaptive occupancy rate, and dynamic interference compensation coefficient of the beacon node are calculated to generate a three-dimensional feature vector. Specifically, the following steps are performed: The core features of the spatiotemporal dynamic feature matrix, such as the signal strength mean, variance, and number of cluster coverage nodes, are extracted by node ID, frequency band, and time window. The high-dimensional features are reduced using principal component analysis to generate a spatiotemporal feature vector. The spatiotemporal feature vector, adaptive occupancy rate, and dynamic interference compensation coefficient are concatenated in a fixed order to form a three-dimensional feature vector that includes spatiotemporal distribution, occupancy efficiency, and interference impact.

[0098] 96 Performing dimensionality reduction on the principal components of the three-dimensional eigenvector to obtain a triplet of characteristic parameters including frequency band activity, stability, and interference sensitivity, specifically comprising: normalizing the data of each dimension of the three-dimensional eigenvector, calculating the covariance matrix of the normalized data of each dimension, performing eigenvalue decomposition (in the eigenvector obtained by eigenvalue decomposition, each element corresponds to the linear correlation coefficient between the original feature and the principal component, i.e., the principal component load weight), obtaining the principal components and their corresponding eigenvalues, sorting them from large to small by eigenvalue, selecting the principal components with a cumulative contribution rate of more than 85%, combining the number of cluster coverage nodes and the mean signal strength with the adaptive occupancy rate in the spatiotemporal eigenvector to map them to frequency band activity, mapping features such as signal strength variance and multipath fading factor that characterize signal fluctuation and transmission stability to stability, mapping features affected by beacon node interference in the dynamic interference compensation coefficient and Doppler frequency shift rate to interference sensitivity, allocating the contributions of the frequency band activity parameter, stability parameter, and interference sensitivity parameter according to the principal component load weight, and generating an ordered triplet of characteristic parameters including frequency band activity, stability, and interference sensitivity;

[0099] Dynamically adjust the feature screening threshold according to the current network load, calculate the local density and distance of the feature parameters in the feature parameter triplet, and generate a frequency band feature parameter set. Specifically, according to the current network load (real-time throughput of beacon nodes and number of concurrent connections), use a dynamic threshold algorithm to adjust the feature screening standard: when the load is high, set the stability threshold (calculate the 90th percentile as the threshold through historical data), set a lower tolerance value for interference sensitivity (take 80% of the recent minimum value as the tolerance value), and use a loose threshold (average value) when the load is low. Standard deviation), after standardizing the characteristic parameter triples, the density reachable algorithm is used to calculate the local density: the dynamic radius (dynamic radius Initial value (1 Current load The neighborhood is delineated by the maximum load) and the number of points in the dynamic radius neighborhood of each frequency band in three-dimensional space is used as the local density. The characteristic distance between frequency bands is calculated by the Manhattan distance algorithm. Load-related density threshold, stability Dynamic noise floor and interference sensitivity The load adjustment coefficient filters effective features, and integrates the filtered triplets with the statistics of local density and average distance to generate a set of frequency band feature parameters including the frequency band activity status, stability characteristics and interference response.

[0100] By proposing a dual calculation method of signal quality factor and adaptive occupancy rate, the problem of inconsistent spectrum data fusion in traditional monitoring systems in complex scenarios is effectively solved. The calculation formula of the signal quality factor integrates multi-dimensional parameters such as standard deviation, mean, and Doppler frequency shift rate, so that spectrum quality assessment is no longer limited to a single signal strength indicator, and can more comprehensively reflect the transmission stability and anti-interference capability of the channel. Secondly, the weighted calculation method of adaptive occupancy rate dynamically adjusts the weight of the frequency band occupancy time according to the signal quality, avoiding the deviation caused by the traditional fixed time statistical method in blind spots or data conflict areas. This enables the system to accurately capture sudden changes in the local spectrum even when the monitoring nodes are unevenly distributed or there are blind spots, thereby improving the perception sensitivity of spectrum anomalies.

[0101] By constructing characteristic parameter triplets and frequency band characteristic parameter sets, the intelligence level and dynamic adaptability of spectrum management are significantly enhanced. On the one hand, the characteristic parameter triplets organically integrate spatiotemporal characteristics, interference sensitivity and stability, and achieve efficient compression and accurate characterization of spectrum characteristics through principal component analysis dimensionality reduction and dynamic threshold screening. On the other hand, the mechanism of dynamically adjusting the screening criteria based on network load enables the system to give priority to retaining high-quality frequency band characteristics with high stability and low interference sensitivity under high load, and relax the screening conditions under low load to explore potential available frequency bands. The adaptive spectrum feature extraction method not only effectively addresses the consistency problem during distributed node data fusion, but also can flexibly adjust the spectrum monitoring strategy according to the load requirements of different scenarios, optimize spectrum resource allocation, and improve the overall performance and reliability of wireless communication networks in complex dynamic environments.

[0102] In one case of this embodiment, adjusting the monitoring range of a common node according to an effective coverage radius parameter includes:

[0103] Based on the effective coverage radius parameters of the beacon node and the overlap requirements of the monitoring area, the monitoring range adjustment coefficient of the ordinary node is calculated, specifically including: determining the three-dimensional coordinates of the beacon node and the ordinary node, calculating the straight-line distance between adjacent beacon nodes, taking the beacon node coverage radius as the benchmark, combined with the overlap requirements (the overlapping area of adjacent monitoring areas is not less than 30% of the single node coverage area as the overlap threshold), deriving the theoretical boundary of the monitoring range of the ordinary node through the geometric projection algorithm, using the grid division method to discretize the monitoring area into grid units, counting the number of times each grid is covered by the beacon node and the ordinary node, and constructing the coverage density matrix. When the coverage density of the monitoring area is lower than the overlap threshold, the monitoring radius of the ordinary node is proportionally expanded, otherwise it is reduced. Through iterative calculation, the monitoring range of each ordinary node is effectively connected with the coverage area of the beacon node in accordance with the overlap requirements. Finally, the ratio of the adjusted monitoring radius to the initial radius is used as the monitoring range adjustment coefficient of the ordinary node.

[0104] The original monitoring range of the ordinary node is combined with the monitoring range adjustment coefficient to generate a new monitoring range, specifically including: multiplying the initial monitoring range of the ordinary node by the monitoring range adjustment coefficient to obtain the adjusted monitoring radius; through the three-dimensional space buffer analysis algorithm, with the coordinates of the ordinary node as the center and the adjusted monitoring radius as the new radius, a spherical basic monitoring area is generated, which is the new monitoring range.

[0105] By dynamically adjusting the monitoring range of ordinary nodes, the problem of distributed node monitoring blind spots and data fusion inconsistency in complex urban activity scenarios is effectively solved. First, based on the monitoring range adjustment mechanism of the effective coverage radius and overlap requirements of the beacon node, the geometric projection and grid division algorithm are used to accurately deduce the monitoring boundary of ordinary nodes and construct a coverage density matrix. It can achieve adaptive optimization of the monitoring area in the iterative process, ensuring that the overlap of the monitoring areas of adjacent nodes is not less than 30%, eliminating the monitoring blind spots from the root. Secondly, the monitoring range is regenerated through the three-dimensional spatial buffer analysis algorithm, which realizes the dynamic reconstruction of the monitoring area and enhances the ability of multi-node collaborative monitoring. It not only significantly improves the monitoring sensitivity of local area spectrum changes, enabling the system to detect spectrum anomalies more timely and accurately, but also provides more accurate spectrum usage data for subsequent channel allocation, effectively avoiding the problem of unreasonable channel allocation due to missing or conflicting monitoring data, thereby significantly improving the reliability and efficiency of the distributed spectrum monitoring system in complex dynamic environments.

[0106] In one case of this embodiment, a frequency band characteristic parameter set is extracted to generate a shared parameter set including a spectrum hole location and occupancy, including:

[0107] According to the new monitoring range, the signal strength and occupancy time data of each frequency band in the dynamic monitoring area are extracted from the frequency band characteristic parameter set to generate spatiotemporal feature association parameters, specifically including: according to the new monitoring range, the node ID falling into the monitoring area is extracted by the spatial screening algorithm, the signal strength mean and variance of the corresponding node are screened from the spatiotemporal dynamic feature matrix of the frequency band characteristic parameter set by frequency band and timestamp index, the occupancy time data required for adaptive occupancy calculation are simultaneously extracted, the data is divided into fixed time windows, the data in each window is spatiotemporally aligned, and the spatiotemporal feature association parameters containing signal strength statistics and weighted occupancy time are generated with node ID, frequency band and timestamp as keys;

[0108] Based on the spatiotemporal feature correlation parameters, the signal strength threshold and the occupancy time threshold are set, and the areas where the signal strength is lower than the signal strength threshold and the occupancy time is lower than the occupancy time threshold are screened out to obtain the spectrum hole position parameters. Specifically, the 10th percentile of the signal strength of each frequency band is calculated as the set signal strength threshold, 30% of the historical average occupancy time is set as the occupancy time threshold, and the signal strength mean and weighted occupancy time data of each node ID, frequency band, and timestamp are subjected to dual-condition screening, that is, the signal strength mean Signal strength threshold and weighted occupancy time The occupation time threshold is used to extract the geographical coordinates and corresponding frequency bands of nodes that meet the conditions. The continuous area of the spectrum hole is fitted through the spatial interpolation algorithm to generate spectrum hole location parameters including coordinate position, frequency band and idle state.

[0109] For the monitoring area corresponding to the spectrum hole location parameter, the proportion of the occupancy time of each frequency band in the monitoring area to the total monitoring time is calculated to obtain the spectrum hole occupancy rate parameter. Specifically, the following steps are performed: spatially match the node geographic coordinates with the continuous area of the spectrum hole through a spatial overlay analysis algorithm, extract the node IDs and corresponding frequency bands falling within the area, group them by frequency band, summarize the weighted occupancy time of each frequency band in all time windows, and calculate the total occupancy time of the frequency band in the monitoring area. The total monitoring time is the total duration of the same period of the spatiotemporal dynamic feature matrix (that is, the sum of the time of all time windows). Using a data aggregation algorithm, the total occupancy time of each frequency band is divided by the total monitoring time to obtain the normalized spectrum hole occupancy rate parameter.

[0110] The spectrum hole location parameters and the spectrum hole occupancy parameters are integrated to generate a shared parameter set containing the spectrum hole location and occupancy, specifically including: based on the coordinate position, frequency band and idle status in the spectrum hole location parameters and the frequency band and occupancy values in the spectrum hole occupancy parameters, the fields are associated through a spatial connection algorithm, and the frequency band and geographic coordinates are used as common indexes. The coordinate position, frequency band number, idle status and corresponding occupancy parameters of each spectrum hole area are attribute-integrated to construct a structured data set containing the fields of frequency band ID, geographic coordinates, idle status and occupancy, and are stored in a unified data format to generate a shared parameter set containing the spectrum hole location and occupancy.

[0111] By precisely locating spectrum holes and determining their occupancy, the system effectively addresses the blind spot problem of spectrum resource monitoring in complex scenarios. By extracting key data for each frequency band within the dynamic monitoring area from a set of frequency band characteristic parameters and generating spatiotemporal characteristic correlation parameters, the system achieves a comprehensive and detailed understanding of spectrum usage. Furthermore, by setting reasonable signal strength and occupancy duration thresholds, the system identifies the true locations of spectrum holes and uses a spatial interpolation algorithm to fit their continuous regions, making the location of spectrum holes more accurate and intuitive. This not only improves spectrum resource utilization but also provides precise idle frequency band information for subsequent channel allocation, avoiding irrational channel allocation caused by inaccurate monitoring, thereby improving the performance and efficiency of the entire communication system.

[0112] By generating a shared parameter set containing spectrum hole locations and occupancy rates, the data collaboration between distributed monitoring nodes is greatly enhanced. On the one hand, by integrating spectrum hole location parameters and occupancy rate parameters to construct a structured shared parameter set, unified management and efficient sharing of data from different nodes are achieved, eliminating inconsistencies during data fusion. On the other hand, the use of a unified data format for storage facilitates rapid exchange and reading of spectrum hole information between nodes, enabling the system to promptly and accurately detect drastic changes in spectrum usage in local areas, and improving its sensitivity to spectrum anomalies. In complex large-scale urban activity scenarios, this efficient collaborative mechanism ensures the dynamic allocation and rational utilization of spectrum resources, providing strong support for addressing the challenges of tight spectrum resources and diversified communication needs.

[0113] In one case of this embodiment, the channel allocation benefit parameters of the shared parameter set are calculated to obtain the final channel allocation strategy parameters of each game area, including:

[0114] Based on the spectrum hole location and occupancy in the shared parameter set, the availability of each frequency band is determined, and a frequency band availability parameter is generated, specifically including: based on the shared parameter set, grouping by frequency band, correlating the geographic coordinates, frequency band and idle status of the spectrum hole location parameter with the frequency band and occupancy value of the occupancy parameter through a spatial connection algorithm, for each frequency band, extracting the coordinates and occupancy of the continuous area corresponding to the spectrum hole, using a data aggregation algorithm to calculate the average occupancy of the frequency band in the monitoring area, and setting an occupancy threshold in combination with the idle status; if the occupancy is lower than the threshold and there is a continuous hole area, then the frequency band is determined to be available in the corresponding area, and a frequency band availability parameter including the frequency band ID, available area and availability status is generated;

[0115] 113 The frequency band availability parameters are combined with the spectrum requirements of each gaming area and analyzed to generate a preliminary channel allocation strategy. Specifically, based on the frequency band availability parameters, a spatial overlay analysis algorithm is used to spatially match the geographical scope of each gaming area with the frequency band available area, extract the available area range and status of each frequency band in each gaming area, and combine the spectrum requirements of the gaming area (target frequency band list and minimum bandwidth requirements). A multi-attribute decision-making algorithm is used to evaluate frequency band occupancy, signal stability, and interference sensitivity to calculate a comprehensive priority score for each available frequency band. After sorting by score from high to low, a greedy allocation strategy is used to prioritize high-priority frequency bands for gaming areas with urgent spectrum needs. At the same time, a neighborhood conflict detection algorithm is used to avoid co-frequency interference between adjacent gaming areas (i.e., checking whether adjacent units have been allocated the same frequency band, and if so, selecting the suboptimal frequency band). Finally, a preliminary channel allocation strategy is generated that includes the frequency band ID, allocated gaming area, available time period, and power limit.

[0116] 114 Calculate the benefits of the preliminary channel allocation strategy to obtain the channel allocation benefit parameters. Specifically, based on the preliminary channel allocation strategy, extract the three indicators of spectrum hole occupancy, signal stability, and interference sensitivity from the shared parameter set for each game area and its allocated frequency band. Simultaneously quantify the spectrum demand score of the game area, specifically: target frequency band matching degree (1 if the allocated frequency band is within the frequency band of the game area, otherwise 0) Bandwidth satisfaction (ratio of available bandwidth of the allocated frequency band to the minimum demand of the gaming area, normalized to 0) 1) Weighted calculations are performed on occupancy (weight 0.3), stability (weight 0.4), interference sensitivity (weight 0.2), and demand satisfaction (weight 0.1) to obtain the basic revenue value of each gaming area. The frequency bands allocated to adjacent gaming areas are compared using a neighborhood conflict detection algorithm. If the frequency bands are the same, the signal overlap area is identified based on the effective coverage radius parameter of the beacon node, combined with a spatial screening algorithm. The revenue of the affected area is deducted based on the area ratio (if the overlap rate is greater than 10%, 20% is deducted, and the ratio is 1:2, and so on). Finally, the regional revenue is summarized according to the load weight of each gaming area (the proportion of real-time throughput of the beacon node) to obtain the total revenue. This generates a channel allocation revenue parameter that includes the total revenue, the contribution value of each indicator (spectrum hole occupancy, signal stability, interference sensitivity, spectrum demand score), and the basic revenue value of each gaming area.

[0117] The channel allocation benefit parameters are calculated to obtain the final channel allocation strategy parameters of each game area, specifically including: based on the channel allocation benefit parameters, with the dual objectives of maximizing the basic benefit value of each game area and minimizing the neighborhood conflict deduction, a multi-objective optimization mechanism is introduced, specifically: combining the frequency band availability parameters (including available area, occupancy threshold, signal stability index) and spectrum demand constraints (target frequency band list matching, allocated bandwidth ≥ minimum required bandwidth), the frequency band allocation in the preliminary strategy is evaluated region by region, for adjacent areas with co-frequency interference, the signal overlapping area is accurately located through the neighborhood conflict detection algorithm, the overlapping area ratio is calculated based on the beacon node effective coverage radius parameter, and the conflict level is quantified (if the overlap rate is greater than 10%, 20% will be deducted, according to the ratio of 1:2, and so on). During the adjustment process, for the game area that needs to reallocate the frequency band, from its available frequency band set, the multi-attribute decision algorithm is used to screen the target frequency bands included in the regional demand list and the available bandwidth The suboptimal frequency band with the minimum required bandwidth gives priority to protecting the basic revenue of the gaming area with urgent spectrum demand, and updates the allocation plan in rounds through iterative dynamic allocation: after each round of adjustment, recalculate the basic revenue, conflict deduction value and total revenue of each area until the total revenue fluctuation range of two adjacent iterations is lower than the convergence threshold (total revenue change rate <5%), and all allocated frequency bands meet the target frequency band matching Bandwidth requirements The three constraints of no adjacent co-channel interference are finally used to generate the final channel allocation strategy parameters.

[0118] Through precise frequency band availability analysis and multi-dimensional channel allocation strategy generation, the problem of unreasonable channel allocation in complex scenarios is effectively solved. Based on the spectrum hole location and occupancy rate in the shared parameter set, the availability of each frequency band can be accurately determined, and detailed frequency band availability parameters can be generated. This step realizes the refined evaluation of frequency band availability through spatial connection algorithm and data aggregation algorithm, ensuring the scientificity and accuracy of frequency band allocation. Combined with the spectrum requirements of each game area, a multi-attribute decision algorithm and greedy allocation strategy are adopted to generate a preliminary channel allocation strategy, which fully considers multiple key factors such as frequency band occupancy rate, signal stability, and interference sensitivity, improves the rationality and efficiency of channel allocation, and effectively avoids co-frequency interference between adjacent game areas through the neighborhood conflict detection algorithm, further improving the reliability of channel allocation. Ultimately, the generated channel allocation strategy can better meet the spectrum requirements of each game area, improve the utilization rate of spectrum resources, and provide a strong guarantee for the stable operation of the communication system.

[0119] Through comprehensive channel allocation benefit calculation and a multi-objective optimization mechanism, the overall benefits of channel allocation are significantly improved. Based on the preliminary channel allocation strategy, by quantifying indicators such as spectrum hole occupancy, signal stability, and interference sensitivity, combined with the spectrum demand score, the basic benefit value of each game area can be accurately calculated. At the same time, the neighborhood conflict detection algorithm and the beacon node effective coverage radius parameter are used to reasonably deduct the benefits of co-frequency interference areas, making the total benefit calculation more reasonable. By introducing a multi-objective optimization mechanism, with the optimization goals of maximizing the basic benefit value of each game area and minimizing neighborhood conflict deductions, the allocation scheme is updated round by round through iterative dynamic allocation until the convergence conditions and constraints are met. This process not only improves the global optimality of channel allocation, but also ensures efficient utilization of spectrum resources and stable operation of the system. The final channel allocation strategy parameters can better adapt to the changes in spectrum demand in complex dynamic environments, providing an efficient and reliable channel allocation solution for communication systems in large-scale urban activity scenarios, significantly improving the overall system performance and user experience.

[0120] In one scenario of this embodiment, the real-time spectrum monitoring data of the monitoring nodes in each game area are subjected to spatiotemporal alignment processing according to the final channel allocation strategy parameters to generate global spectrum state parameters including signal strength gradient distribution and spectrum state mutation probability, including:

[0121] Based on the time slice division in the final channel allocation strategy parameters, an adaptive sliding time window is generated for each monitoring node. The signal propagation delay is calculated and the time window boundary is dynamically adjusted through the time stamp signal emitted by the beacon node and the received signal strength of the ordinary node to generate the time-space synchronization calibration parameters. Specifically, it includes: based on the time slice division of the final channel allocation strategy, an adaptive sliding time window is generated for each monitoring node, wherein the beacon node periodically transmits the time stamp signal containing the time stamp, and the ordinary node receives and records the signal strength and arrival time. A two-way timestamp algorithm is used to calculate the signal propagation delay through the sending and receiving time difference of the signal interaction between the beacon node and the ordinary node. The time window boundary of the ordinary node is dynamically adjusted according to the signal propagation delay. With the beacon node time slice as the benchmark, the delay is compensated to the start and end time of the ordinary node time window, so that the time window slides adaptively with the propagation delay, and the time-space synchronization calibration parameters including the node time synchronization deviation are generated;

[0122] The weight index of common monitoring nodes is dynamically adjusted according to the distribution density of monitoring nodes and the effective coverage radius parameters of beacon nodes to generate signal strength gradient distribution parameters. Specifically, the spatial distribution density of common nodes in the monitoring area is calculated by the kernel density estimation algorithm, a buffer zone is constructed with the beacon node as the center, and the initial weight index is generated by the inverse distance weighting algorithm according to the ratio of the Euclidean distance from the common node to the beacon node to the coverage radius. The weight index of nodes with a distribution density higher than the threshold (the historical density mean) is increased. For areas with large standard deviations, the weight is attenuated according to the density ratio. For low-density areas, the weight is compensated according to the coverage radius. Finally, the signal strength of common nodes is spatially weighted interpolated with the adjusted weight index to generate the signal strength gradient distribution parameter.

[0123] The spatiotemporal synchronization calibration parameters and the spectrum state differences of adjacent time slices are analyzed to obtain the probability of spectrum state mutation. The specific process is: for each time slice , will The signal strength change of each time slice is divided by the reference signal strength benchmark and the absolute value is taken. The power operation is performed to obtain the nonlinear amplification value of the signal intensity change of the mth time slice, and the amplification value is multiplied by the sign function. If ,but ,if ,but ,if ,but , then sum the results of all time slices to get the comprehensive mutation trend value, sum the amplification values of all time slice signals to get the sum of the nonlinear amplification values of the signal intensity changes of all time slices, take the sum and the zero elimination constant The maximum value among them is multiplied by the environmental attenuation factor , get the total fluctuation intensity value after environment adjustment, divide the comprehensive mutation trend value by the total fluctuation intensity value after environment adjustment to get the spectrum state mutation probability;

[0124] When applied in practice, the specific calculation formula for the above spectrum state mutation probability is as follows:

[0125] ;

[0126] Where, represents the probability of spectrum state mutation, Indicates the The signal strength change of each time slice is Indicates the reference signal strength benchmark, represents the nonlinear amplification exponent, , represents the environmental attenuation factor, , Indicates the zero-proof constant, , represents the symbolic function, for , is the total number of time slices;

[0127] The signal strength gradient distribution parameters and the probability of spectrum state mutation are analyzed to generate global spectrum state parameters including the signal strength gradient distribution and the probability of spectrum state mutation. Specifically, based on the signal strength gradient distribution parameters, the monitoring area is discretized into regular grid cells through a spatial gridding algorithm, the signal strength gradient value of each grid cell is extracted, and the spectrum state mutation probability of each grid cell in adjacent time slices is synchronously calculated. The gradient distribution value of the grid cell is matched with the mutation probability of the corresponding position using a spatiotemporal data association algorithm. A multidimensional data set including spatial gradient characteristics and temporal mutation characteristics is constructed with node ID, frequency band and timestamp as indexes. After integration in a unified data format, global spectrum state parameters including the signal strength gradient distribution and the probability of spectrum state mutation are generated.

[0128] Through the final channel allocation strategy, an adaptive sliding time window is generated for each monitoring node, and the signal propagation delay is calculated using a two-way timestamp algorithm to dynamically adjust the time window boundary and generate spatiotemporal synchronization calibration parameters, thereby achieving spatiotemporal synchronization between monitoring nodes and enabling more accurate monitoring of the spectrum status. On the other hand, the weight index of ordinary monitoring nodes is dynamically adjusted by comprehensively considering factors such as the distribution density of monitoring nodes and the effective coverage radius of beacon nodes, and the signal strength gradient distribution parameters are generated, making the spectrum monitoring data more representative in spatial distribution. Even in large-scale urban activity scenarios with uneven node distribution, it can effectively reduce blind spots and improve the monitoring accuracy of spectrum usage, thereby providing a more reliable data basis for subsequent timely detection of spectrum anomalies and reasonable channel allocation.

[0129] By conducting an in-depth analysis of the spatiotemporal synchronization calibration parameters and the spectrum state differences between adjacent time slices, the probability of spectrum state mutation is derived and combined with the signal strength gradient distribution parameters to generate global spectrum state parameters. This significantly improves the monitoring sensitivity to sharp changes in spectrum usage in local areas. Nonlinear amplification and sign function methods are used to process the signal strength changes in each time slice, effectively amplifying the mutation trend of the spectrum state. Even in the early stages when the spectrum changes are relatively subtle, they can be detected in a timely manner, avoiding the opportunity to discover spectrum anomalies due to low monitoring sensitivity. Then, based on the signal strength gradient distribution parameters, a spatial gridding algorithm and a spatiotemporal data association algorithm are used to combine spatial gradient characteristics with temporal mutation characteristics to construct a multidimensional dataset. This can more accurately present the dynamic changes of the spectrum and provide a more comprehensive and detailed decision-making basis for channel allocation, effectively avoiding the problem of irrational channel allocation caused by the failure to detect spectrum anomalies in a timely manner, and ensuring the efficient utilization of spectrum resources in large-scale urban activity scenarios.

[0130] In one case of this embodiment, judging whether the current channel allocation strategy is invalid based on the global spectrum status parameter, and immediately triggering the dynamic allocation instruction if invalid, includes:

[0131] The historical spectrum monitoring data is divided into k time slices and frequency bands, and the gradient mean of the spatial distribution of the signal strength of each frequency band and the change rate of the spectrum occupancy in the time dimension are calculated slice by slice, forming a dynamic feature sequence that contains the spatiotemporal variation law. Specifically, the historical spectrum monitoring data is divided into k time slices according to the minute-level time granularity, and the frequency domain is divided according to the MHz-level frequency band accuracy. For each frequency band in each time slice, the signal strength data of discrete monitoring nodes are spatially interpolated using the Kriging interpolation algorithm based on the geographical coordinates and signal strength data of the monitoring nodes to obtain a continuous signal strength spatial distribution. The signal strength gradient value of each spatial position is calculated using the gradient operator, and the average of the gradient values in the region is taken as the signal strength gradient mean of the frequency band and time slice. For the spectrum occupancy sequence in the time dimension of each frequency band, the absolute change value of the occupancy of adjacent time slices is calculated using the first-order difference method, and the absolute change value is divided by the time slice length to obtain the time dimension change rate, which is the mutation probability. The gradient mean and change rate are indexed by the time slice-frequency band to form a dynamic feature sequence that contains the spatiotemporal variation law.

[0132] The time period data corresponding to the historical channel allocation are grouped by the game area, and the median of the signal strength gradient and the extreme value of the spectrum occupancy change rate in the dynamic feature sequence of each monitoring area are counted respectively to generate preliminary regional gradient benchmark parameters and regional mutation benchmark parameters. Specifically, according to the geographical scope of the game area, the historical channel allocation time period data are spatially matched with the game area polygon through the spatial connection algorithm, and the dynamic feature sequence of the corresponding time period in each area is first extracted. For each game area, the dimensions are grouped according to the time slice-frequency band, and the median of the signal strength gradient mean within the group is calculated. At the same time, the maximum and minimum values of the spectrum occupancy change rate (the extreme value of the mutation amplitude in the time dimension) are extracted. The median is used as the preliminary regional gradient benchmark parameter, and the extreme value of the mutation amplitude is used as the preliminary regional mutation benchmark parameter.

[0133] Based on the building density of the current monitoring area, the preliminary regional gradient benchmark parameters and regional mutation benchmark parameters are corrected to obtain the final signal strength gradient threshold and mutation probability threshold that are suitable for the current monitoring environment. Specifically, based on the building density grid data of the monitoring area, the average building density value of each game area is obtained through the partition statistical algorithm, and the area is divided into three levels of high density, medium density and low density according to the density. The ratio of the median gradient to the average density in the historical data of each level area is statistically used as the correction coefficient, and the median gradient of the corresponding level is multiplied by the correction coefficient to obtain the signal strength gradient threshold that is suitable for the current density. For the extreme value of the spectrum occupancy rate change, according to the density grading result, the average attenuation amplitude of the mutation extreme value under each density level is statistically calculated using historical data, and the original extreme value is proportionally adjusted in the same direction through the hierarchical weighted algorithm to generate the final mutation probability threshold, and the final signal strength gradient threshold and mutation probability threshold that are suitable for the current monitoring environment are obtained.

[0134] By constructing a dynamic feature sequence that incorporates the laws of spatiotemporal variation, the accuracy and sensitivity of spectrum status monitoring in complex large-scale urban activity scenarios are effectively improved. Historical spectrum monitoring data is finely segmented by time slices and frequency bands, and the Kriging interpolation algorithm is used to achieve spatial continuity of signal strength. Combined with the gradient operator and the first-order difference method, the dynamic characteristics of each frequency band in terms of spatial gradient and temporal change rate are accurately extracted. This process not only captures subtle changes in spectrum status, but also forms a dynamic feature sequence with spatiotemporal dimensions, providing high-resolution data support for subsequent channel allocation strategy evaluation. In large-scale urban activity scenarios, accurate dynamic feature extraction can effectively compensate for the blind spots of distributed monitoring nodes and avoid monitoring deviations caused by insufficient data fusion and coordination mechanisms, thereby significantly enhancing monitoring sensitivity to rapid changes in spectrum usage. Through the generation of this dynamic feature sequence, the system can more keenly detect spectrum anomalies, facilitating timely adjustment of channel allocation strategies to ensure efficient utilization of spectrum resources.

[0135] By statistically analyzing the dynamic feature sequences in different game areas, preliminary regional gradient benchmark parameters and regional mutation benchmark parameters are obtained. These benchmark parameters are then graded and corrected according to the building density to generate the final signal strength gradient threshold and mutation probability threshold that are adapted to the current environment. This parameter correction method based on environmental density fully considers the impact of buildings on signal propagation in large-scale urban activity scenarios, allowing the channel allocation strategy to be dynamically adjusted according to the actual monitoring environment. In areas with high building density, the system can more accurately identify spectrum state changes through the corrected thresholds, avoiding monitoring errors caused by environmental differences. In medium and low density areas, it can optimize resource allocation and reduce over-monitoring or under-monitoring. Through this precise parameter correction and adaptation, the system can promptly determine whether the current channel allocation strategy has failed and trigger dynamic allocation instructions when necessary, greatly improving the rationality of channel allocation and the efficiency of spectrum management. It effectively solves the problem of unreasonable channel allocation caused by the lack of environmental adaptability in traditional methods, and provides a reliable solution for spectrum resource management in large-scale urban activities.

[0136] In one case of this embodiment, judging whether the current channel allocation strategy is invalid according to the global spectrum status parameter, and immediately triggering the dynamic allocation instruction if invalid, further includes:

[0137] If the signal strength gradient in the global spectrum state parameter is greater than the signal strength gradient threshold or the mutation probability is continuously higher than the mutation probability threshold for k consecutive time slices, the previous channel allocation strategy is determined to be invalid and the dynamic allocation instruction is triggered immediately.

[0138] By setting clear threshold judgment criteria, the problem of difficult timely detection of channel allocation strategy failure in complex large-scale urban activity scenarios is effectively solved. When the signal strength gradient in the global spectrum state parameter exceeds the set threshold, or the mutation probability continues to be higher than the threshold for k consecutive time slices, the system can quickly determine that the current channel allocation strategy has failed and immediately trigger the dynamic allocation instruction, thereby improving the response speed to spectrum state changes and avoiding delayed detection of spectrum anomalies due to insufficient monitoring sensitivity. In large-scale urban activity scenarios, channel allocation can be dynamically adjusted in a timely manner according to the actual use of the spectrum to ensure the efficiency and stability of communication, effectively making up for the defects of unreasonable channel allocation caused by blind spots of monitoring nodes and inconsistent data fusion in traditional methods.

[0139] In one case of this embodiment, the dynamic allocation instruction includes:

[0140] The area where the signal intensity gradient exceeds the signal intensity gradient threshold is marked as an abnormal area, and the spatiotemporal correlation parameters between the abnormal area and the adjacent areas are calculated to generate the regional abnormal propagation factor. Specifically, the signal intensity gradient value of each grid cell is compared with the signal intensity gradient threshold adapted to the current monitoring environment based on the discretized regular grid cells in the global spectrum state parameters. The grid cells whose gradient values exceed the corresponding threshold constitute abnormal areas. The spatial connection algorithm is used to identify the adjacent areas directly adjacent to the abnormal area in space. For the abnormal area and each adjacent area, the signal intensity gradient change and spectrum state mutation probability data of the two are extracted in the same frequency band dimension and adjacent time slice sequence. The spatiotemporal correlation parameters are obtained by calculating the correlation coefficient of the two sets of data. The spatial weight matrix is constructed based on the Euclidean distance between each adjacent area and the abnormal area. The spatiotemporal correlation parameters and the spatial weight are weightedly fused to generate the regional abnormal propagation factor.

[0141] The signal strength gradient distribution of each monitoring node in the abnormal area is analyzed to obtain the spectrum state chaos degree. Specifically, the following steps are performed: for each monitoring node in the abnormal area, the signal strength gradient value corresponding to its geographic coordinates is extracted to form a two-dimensional data set containing location information and gradient values. The gradient values are normalized, and the distribution of the gradient values of all nodes is statistically analyzed. The frequency of occurrence of each gradient value in the area (i.e., the proportion of the number of nodes with this gradient value to the total number of nodes in the abnormal area) is calculated to generate a discrete probability density distribution. For each unique gradient value, its frequency of occurrence is regarded as a probability. The sum of the products of all gradient value probabilities and the corresponding natural logarithms is calculated, and the negative value is taken. This value is the spectrum state chaos degree.

[0142] Based on the regional anomaly propagation factor and spectrum state disorder, the boundaries of the game area are dynamically adjusted to generate a temporary spectrum management area. Specifically, a spatial impact buffer is constructed based on the numerical value of the regional anomaly propagation factor, with the geometric center of the anomaly area as the benchmark. A higher propagation factor indicates a greater potential for the anomaly area to have a spatiotemporal impact on the surrounding area. The buffer radius is dynamically expanded by a proportional coefficient (each unit propagation factor corresponds to a 150-meter buffer distance). A spatial adjacency network between the anomaly area and all adjacent areas is constructed using the Delaunay triangulation algorithm, and directly adjacent polygonal areas are extracted. The spectrum state disorder is used as a criterion for regional stability. For anomaly areas with a disorder degree exceeding 1.5 standard deviations from the historical mean, an adjacent area merging mechanism is activated. Specifically, for each adjacent area, a weighted composite value of its propagation factor and the disorder degree of the anomaly area is calculated (the propagation factor accounts for 60% and the disorder degree accounts for 40%). When the composite value exceeds a dynamic threshold (0.65), the adjacent area is included in the temporary management scope using a polygon overlay algorithm. Based on the original boundary of the anomaly area, the polygons of the adjacent areas that meet the conditions are merged to generate a temporary spectrum management area encompassing the anomaly impact range.

[0143] The monitoring nodes in the temporary spectrum management area are analyzed to obtain the channel resource desirability vector, which specifically includes: extracting the signal strength, spectrum occupancy and number of channel conflicts of each node in the current and historical time slices for the monitoring nodes in the temporary spectrum management area, calculating the spectrum occupancy change rate of the node in the target frequency band (the difference in occupancy of adjacent time slices divided by the length of the time slice) as the demand urgency indicator, counting the number of channel retransmissions of the node in the past 30 minutes as the congestion sensitivity indicator, setting the priority weight according to the node service type (real-time service empowerment 0.6, non-real-time service empowerment 0.4), normalizing the above indicators, and constructing a multidimensional vector according to the frequency band dimension (10 MHz frequency band), and the component corresponding to the position of each frequency band is the occupancy change rate Priority Weight Congestion sensitivity (1 Priority weight), forming a channel resource desirability vector that includes frequency band priority and demand intensity;

[0144] The regional abnormal propagation factor, spectrum state chaos, and channel resource desirability vector are extracted to obtain multidimensional risk assessment parameters. This includes: cross-dimensional feature fusion of the regional abnormal propagation factor, spectrum state chaos, and channel resource desirability vector. Specifically, the propagation factor and chaos are normalized using the Z-score normalization method. The demand components of each frequency band in the channel resource desirability vector are arranged in sequence according to the frequency band order (where 0-99MHz is divided into 10 10MHz frequency bands). A multidimensional vector is constructed in a fixed dimensional order: the first dimension is the normalized regional abnormal propagation factor, the second dimension is the normalized spectrum state chaos, and the third to 12th dimensions correspond to the channel resource desirability components of the ten frequency bands, respectively. Ultimately, a multidimensional risk assessment parameter is formed that includes spatial propagation risk, regional stability risk, and frequency band resource demand.

[0145] When the evaluation value of any dimension of the multidimensional risk evaluation parameters exceeds the multidimensional risk threshold, a dynamic allocation instruction including the optimal channel combination, power adjustment and time slice allocation strategy is generated;

[0146] The multi-dimensional risk thresholds are as follows:

[0147] Statistically analyze the propagation factor distribution of all adjacent regions under historical normal conditions, calculate their mean and standard deviation, and take the mean + 1.5 standard deviations as the regional abnormal propagation factor threshold;

[0148] Perform kernel density estimation on the chaos data of the historical stable area, use the chaos value corresponding to the probability density peak as the benchmark, and set the chaos value in combination with the business tolerance. 1.2 is used as the spectrum state chaos threshold;

[0149] For each 10MHz frequency band, calculate the mean and 95th percentile of the channel resource desirability component under normal service scenarios over the past seven days. Use the 95th percentile as the channel resource desirability component threshold.

[0150] For frequency bands dominated by real-time services, the threshold for the channel resource desirability component will be lowered by 10%, and for frequency bands not used for real-time services, the threshold for the channel resource desirability component will be raised by 15%.

[0151] The dynamic allocation instructions are as follows:

[0152] The trigger conditions are:

[0153] When the regional abnormal propagation factor Regional anomaly propagation factor threshold or when the spectrum state is chaotic The spectrum state chaos threshold or the channel resource demand component of any frequency band The desirability threshold of the corresponding frequency band;

[0154] The optimal channel combination strategy is:

[0155] When the channel resource demand component of any frequency band The frequency bands that exceed the channel resource desirability threshold are sorted from high to low by their desirability components, and the top three most desirable frequency bands are prioritized as core allocation frequency bands.

[0156] When the regional abnormal propagation factor Regional abnormal propagation factor threshold, the frequency band and its adjacent One frequency band (a total of three 10MHz bands) is marked as an interference-sensitive band and is temporarily disabled;

[0157] Generate a channel combination that includes high-demand frequency bands (by priority) and non-interference-sensitive frequency bands to ensure that the real-time service frequency band accounts for ≥ 60%.

[0158] The power adjustment strategy is:

[0159] Core area of the abnormal area (within the buffer zone radius):

[0160] When the regional abnormal propagation factor The regional anomaly propagation factor threshold is used to reduce the power of monitoring nodes in the core area (by 10% of the current power) to reduce signal overflow interference;

[0161] When the spectrum state is chaotic The spectrum chaos threshold indicates that the spectrum state in the area is unstable. Power fine-tuning (±5%) is performed on the top 20% of nodes with the highest gradient values to balance the signal strength distribution in the area.

[0162] Adjacent merge area (extension area outside the buffer zone):

[0163] For adjacent areas included in the temporary management scope, the closer to the abnormal center, the greater the power increase (power per 100 meters distance). 2%, with an upper limit of 30%), to enhance the anti-interference capability of edge node signals.

[0164] The time slice allocation strategy is:

[0165] Real-time business nodes:

[0166] For nodes with high demand urgency (high occupancy rate change), short time slices are allocated (time slice length is shortened by 25%), and they are preferentially inserted into the first 50% of the time slice sequence to reduce transmission delay;

[0167] Nodes with high congestion sensitivity (>10 retransmissions in 30 minutes) are allocated an additional redundant time slice to reduce the probability of collision.

[0168] Non-real-time business nodes:

[0169] For nodes with low demand urgency (low occupancy change rate), the time slice length is extended by 20% and scheduled during off-peak hours (the last 30% of the time slice per cycle);

[0170] For low-demand frequency band nodes that are not affected by abnormal propagation, the time slice allocation interval is expanded by 50% to release resources for high-demand nodes.

[0171] Cross-regional collaborative allocation:

[0172] If the adjacent area merging mechanism is triggered (the chaos degree is greater than 1.5 times the standard deviation), time slices are allocated to the nodes in the merged area in descending order of priority from the abnormal core area to the adjacent expansion area, and the time slice resource share of the core area nodes is ≥ 40%;

[0173] When multiple dimensions exceed the standard simultaneously (the regional abnormal propagation factor, spectrum state disorder, and channel resource demand components of any frequency band are all abnormal):

[0174] On the basis of the highly demanded frequency band, the upper and lower adjacent frequencies of the frequency band where the abnormal area is located are additionally excluded to form an isolated protection frequency band;

[0175] Reduce power consumption in the core area by 15% Bidirectional adjustment of the power increase (20%) in the extended area to build a signal strength gradient buffer zone;

[0176] The time slice priority of real-time business nodes is raised to the highest level, and one pre-allocated time slice is added to all nodes in the abnormal area to ensure the return of monitoring data.

[0177] By accurately identifying abnormal areas and deeply analyzing their propagation characteristics and the degree of spectrum state chaos, it provides strong support for the dynamic adjustment of spectrum management strategies. Based on the global spectrum state parameters, it can accurately mark abnormal areas where the signal strength gradient exceeds the threshold, and use the spatiotemporal correlation parameters to generate regional abnormal propagation factors, clearly quantifying the potential impact range and intensity of the abnormal area on the surrounding area. At the same time, the signal strength gradient distribution of each monitoring node in the abnormal area is analyzed, and the resulting spectrum state chaos effectively reflects the degree of disorder in the use of spectrum in the area. Based on these two key indicators, the boundaries of the game area are dynamically adjusted to generate temporary spectrum management areas, realizing rapid containment and precise control of abnormal situations, significantly improving the system's monitoring sensitivity and response capabilities to local spectrum anomalies, and effectively solving the problem of delayed detection of spectrum anomalies caused by monitoring blind spots and inconsistent data fusion in complex large-scale urban activity scenarios, facilitating the subsequent reasonable allocation of channel resources and ensuring smooth communication.

[0178] By meticulously analyzing monitoring nodes within the temporary spectrum management area, a channel resource desirability vector is generated and combined with multidimensional risk assessment parameters to optimize the rationality of channel allocation decisions. The channel resource desirability vector integrates multidimensional information such as the node's spectrum occupancy rate change, the number of channel conflicts, and service type priority, accurately quantifying the intensity of resource demand in each frequency band. This allows the system to clearly identify which frequency bands are resource-constrained and relatively idle. Furthermore, the system integrates the regional anomaly propagation factor, the spectrum state disorder, and the channel resource desirability vector to form a multidimensional risk assessment parameter, comprehensively considering spatial propagation risk, regional stability risk, and resource demand factors. When the assessment value of any dimension exceeds a threshold, the system rapidly generates dynamic allocation instructions, including the optimal channel combination, power adjustment, and time-slice allocation strategy, achieving refined and dynamic management of channel resources. This dynamic allocation mechanism, based on multidimensional risk assessment, effectively avoids the irrational channel allocation issues inherent in traditional methods, which often result from the inability to accurately detect spectrum anomalies in a timely manner. It significantly improves spectrum resource utilization and ensures the efficient, stable, and reliable operation of communication systems in complex, large-scale urban activity scenarios.

[0179] While embodiments of the present invention have been shown and described, it will be appreciated by those skilled in the art that various changes, modifications, substitutions, and variations may be made to these embodiments without departing from the principles and spirit of the invention, and that the scope of the invention is defined by the appended claims and their equivalents.

Claims

1. A broadband spectrum monitoring system based on dynamic channel allocation, characterized in that: include: A data acquisition unit is used to establish multiple monitoring nodes in the city, divide the city into multiple monitoring areas, and divide the monitoring nodes into beacon nodes and ordinary nodes. The beacon nodes obtain spectrum monitoring data and historical spectrum monitoring data of the monitoring area in real time, dynamically adjust the beacon node's transmission power based on the real-time spectrum monitoring data, synchronously calculate the effective coverage radius parameter of the beacon node, and calculate the signal strength and occupancy time data of each frequency band in the current monitoring area collected in real time by the ordinary nodes to obtain a set of frequency band characteristic parameters; The data identification unit is used to adjust the monitoring range of ordinary nodes according to the effective coverage radius parameter, extract the frequency band characteristic parameter set, and generate a shared parameter set including the location and occupancy of the spectrum hole; A region division unit is configured to divide the monitoring area into multiple game regions, using the beacon nodes in each region as game entities, and calculate the channel allocation benefit parameters of the shared parameter set to obtain the final channel allocation strategy parameters for each game region, including: determining the availability of each frequency band based on the spectrum hole location and occupancy rate in the shared parameter set, and generating a frequency band availability parameter; The frequency band availability parameters are combined with the spectrum requirements of each gaming area and analyzed to generate a preliminary channel allocation strategy; Calculate the benefits of the preliminary channel allocation strategy to obtain the channel allocation benefit parameters; A data alignment unit is used to perform spatiotemporal alignment processing on the real-time spectrum monitoring data of the monitoring nodes in each game area according to the final channel allocation strategy parameters, and generate global spectrum state parameters including signal strength gradient distribution and spectrum state mutation probability; The channel allocation unit is used to determine whether the current channel allocation strategy is invalid based on the global spectrum status parameters, and if invalid, immediately trigger the dynamic allocation instruction; Dynamically adjust the beacon node's transmit power based on real-time spectrum monitoring data, and simultaneously calculate the beacon node's effective coverage radius parameters, including: For each frequency band in the real-time spectrum monitoring data, calculate its and The absolute value of the signal strength difference over time is divided by the dynamic noise floor of the frequency band, and then the sum is calculated for all frequency bands to obtain the total parameter of the change of the signal strength of each frequency band relative to the noise floor. is the length of the dynamic time window; For each frequency band, first calculate the time derivative of the signal strength and multiply it by Then divide it by the mean signal strength of the entire frequency band to obtain the time derivative parameter of the signal strength after frequency band normalization; After squaring the time derivative parameter of the signal strength after frequency band normalization, the sum of all frequency bands is added to 1 to obtain the comprehensive parameter of the relationship between the square of the signal strength change rate and the mean value of all frequency bands; The sum parameter is divided by the comprehensive parameter, and then the result is calculated by the hyperbolic tangent function to obtain the dynamic interference compensation coefficient; The dynamic interference compensation coefficient is combined with the basic transmit power of the beacon node to obtain the current actual transmit power value; The current actual transmission power value, urban building density and wireless propagation loss are integrated and calculated to generate the effective coverage radius parameter.

2. A broadband spectrum monitoring system based on dynamic channel allocation according to claim 1, characterized in that: The signal strength and occupancy duration data of each frequency band in the current monitoring area collected in real time by ordinary nodes are calculated to obtain a set of frequency band characteristic parameters, including: Cluster the signal strength data collected by adjacent common nodes in the same time period to generate a spatiotemporal dynamic feature matrix; By analyzing the spatiotemporal dynamic characteristic matrix, the signal strength fluctuation variance, multipath fading factor and Doppler frequency shift rate of the wireless channel are obtained; Calculate the ratio of the standard deviation to the mean of the signal strength in the frequency band, calculate the ratio of the Doppler frequency shift rate to the maximum Doppler frequency shift rate supported by the system, and multiply the two ratios to obtain the fluctuation-frequency shift coupling factor; Take the square root of the multipath fading factor to get the multipath fading adjustment factor; Divide the fluctuation-frequency shift coupling factor by the multipath fading adjustment factor to obtain the signal dynamic characteristic index; Then calculate the ratio of the maximum and minimum signal strengths of the frequency band in the monitoring area to obtain the signal strength extreme value ratio, and multiply the signal strength extreme value ratio by the signal dynamic characteristic index to obtain the signal quality factor; The traditional occupancy time of a specific frequency band monitored by ordinary nodes is weighted and calculated according to the signal quality factor to generate an adaptive occupancy rate; Calculate the spatiotemporal dynamic feature matrix, adaptive occupancy rate and dynamic interference compensation coefficient of beacon nodes to generate a three-dimensional feature vector; Perform dimensionality reduction on the principal components of the three-dimensional feature vector to obtain a triplet of feature parameters including frequency band activity, stability, and interference sensitivity; The feature screening threshold is dynamically adjusted according to the current network load, the local density and distance of the feature parameters in the feature parameter triplet are calculated, and a frequency band feature parameter set is generated.

3. The broadband spectrum monitoring system based on dynamic channel allocation according to claim 1, characterized in that: Adjust the monitoring range of common nodes based on the effective coverage radius parameters, including: Calculate the monitoring range adjustment coefficient of ordinary nodes based on the effective coverage radius parameters of the beacon node and the monitoring area overlap requirements; The original monitoring range of the ordinary node is combined with the monitoring range adjustment coefficient to generate a new monitoring range, specifically including: multiplying the initial monitoring range of the ordinary node by the monitoring range adjustment coefficient to obtain the adjusted monitoring radius; through the three-dimensional space buffer analysis algorithm, with the coordinates of the ordinary node as the center and the adjusted monitoring radius as the new radius, a spherical basic monitoring area is generated, which is the new monitoring range.

4. The broadband spectrum monitoring system based on dynamic channel allocation according to claim 2, characterized in that: Extract the frequency band characteristic parameter set to generate a shared parameter set containing the spectrum hole location and occupancy, including: According to the new monitoring range, the signal strength and occupancy time data of each frequency band in the dynamic monitoring area are extracted from the frequency band characteristic parameter set to generate spatiotemporal characteristic correlation parameters; Based on the spatiotemporal feature correlation parameters, the signal strength threshold and the occupancy time threshold are set, and the areas with signal strength lower than the signal strength threshold and occupancy time lower than the occupancy time threshold are screened out to obtain the spectrum hole position parameters; Using a spatial overlay analysis algorithm, the geographic coordinates of the nodes are spatially matched with the continuous area of spectrum holes. The node IDs and corresponding frequency bands falling within the area are extracted. The nodes are grouped by frequency bands, and the weighted occupancy time of each frequency band in all time windows is summarized. The total occupancy time of the frequency band in the monitoring area is calculated. The total monitoring time is the total time of the spatiotemporal dynamic feature matrix in the same period. Using a data aggregation algorithm, the total occupancy time of each frequency band is divided by the total monitoring time to obtain the spectrum hole occupancy parameter. The spectrum hole location parameters and the spectrum hole occupancy parameters are integrated to generate a shared parameter set including the spectrum hole location and occupancy.

5. The broadband spectrum monitoring system based on dynamic channel allocation according to claim 1, characterized in that: According to the final channel allocation strategy parameters, the real-time spectrum monitoring data of the monitoring nodes in each game area are temporally and spatially aligned to generate global spectrum state parameters including signal strength gradient distribution and spectrum state mutation probability, including: Based on the time slice division in the final channel allocation strategy parameters, an adaptive sliding time window is generated for each monitoring node. The signal propagation delay is calculated and the time window boundary is dynamically adjusted through the time-standard signal transmitted by the beacon node and the received signal strength of the ordinary node to generate the time-space synchronization calibration parameters. Dynamically adjust the weight index of common monitoring nodes according to the monitoring node distribution density and beacon node effective coverage radius parameters to generate signal strength gradient distribution parameters; The spatiotemporal synchronization calibration parameters and the spectrum state differences of adjacent time slices are analyzed to obtain the probability of spectrum state mutation; The signal intensity gradient distribution parameters and the spectrum state mutation probability are analyzed to generate global spectrum state parameters including the signal intensity gradient distribution and the spectrum state mutation probability.

6. A broadband spectrum monitoring system based on dynamic channel allocation according to claim 5, characterized in that: Based on the global spectrum status parameters, determine whether the current channel allocation strategy is invalid. If invalid, immediately trigger the dynamic allocation instruction, including: The historical spectrum monitoring data is divided into k time slices and frequency bands. The gradient mean of the spatial distribution of the signal strength of each frequency band and the rate of change of the spectrum occupancy in the time dimension are calculated slice by slice, forming a dynamic feature sequence that contains the spatiotemporal variation law. The data of the time period corresponding to the historical channel allocation is grouped by the game area, and the median of the signal strength gradient and the extreme value of the spectrum occupancy change rate in the dynamic feature sequence of each monitoring area are counted respectively to generate preliminary regional gradient benchmark parameters and regional mutation benchmark parameters; Based on the building density of the current monitoring area, the preliminary regional gradient benchmark parameters and regional mutation benchmark parameters are modified to obtain the final signal intensity gradient threshold and mutation probability threshold that are adapted to the current monitoring environment.

7. A broadband spectrum monitoring system based on dynamic channel allocation according to claim 6, characterized in that: Based on the global spectrum status parameters, it is determined whether the current channel allocation strategy is invalid. If invalid, the dynamic allocation instruction is immediately triggered. It also includes: If the signal strength gradient in the global spectrum state parameter is greater than the signal strength gradient threshold or the mutation probability is continuously higher than the mutation probability threshold for k consecutive time slices, the previous channel allocation strategy is determined to be invalid and the dynamic allocation instruction is triggered immediately.

8. The broadband spectrum monitoring system based on dynamic channel allocation according to claim 7, characterized in that: Dynamic dispatch instructions include: The area where the signal intensity gradient exceeds the signal intensity gradient threshold is marked as an abnormal area, and the spatiotemporal correlation parameters between the abnormal area and the adjacent areas are calculated to generate the regional abnormal propagation factor; Analyze the signal strength gradient distribution of each monitoring node in the abnormal area to obtain the spectrum state chaos degree; According to the regional abnormal propagation factor and the disorder degree of the spectrum state, the boundaries of the game area are dynamically adjusted to generate a temporary spectrum management area; Analyze the monitoring nodes within the temporary spectrum management area to obtain the channel resource desirability vector; Extract regional abnormal propagation factors, spectrum state disorder, and channel resource demand vectors to obtain multi-dimensional risk assessment parameters; When the evaluation value of any dimension of the multidimensional risk evaluation parameters exceeds the multidimensional risk threshold, a dynamic allocation instruction including the optimal channel combination, power adjustment and time slice allocation strategy is generated.

Images