A method for wireless communication resource allocation based on dynamic spectrum sensing

Abstract

The application discloses a wireless communication resource allocation method based on dynamic spectrum sensing, which comprises the following steps: acquiring device node data and network environment parameters and preprocessing to generate feature data; based on the preprocessed feature data and time window, a dynamic space-time feature encoding is adopted to generate a space-time feature map to capture the space-time correlation between devices; a hyperedge set is constructed by a hyperedge dynamic generation mechanism using device location information to reflect network topology changes; a high-order interference tensor is constructed based on the hyperedge set, transmission power and channel gain and is decomposed to accurately quantify the interference relationship; according to the interference tensor, the number of resource blocks and the quality of service demand, a spectral clustering and optimization algorithm is adopted to generate an optimal resource allocation strategy; through the space-time attention mechanism and high-order interference tensor modeling, the dynamic adaptability and optimization performance of resource allocation can be improved, and an efficient solution is provided for modern wireless communication networks.

Description

Technical Field

[0001] This invention relates to the field of resource allocation technology in wireless communication, and in particular to a wireless communication resource allocation method based on dynamic spectrum sensing. Background Technology

[0002] Wireless communication resource allocation technology occupies a core position in modern communication systems, and its development is closely related to the efficient utilization of spectrum resources. As mobile communication technology evolves from 4G to 5G and is moving towards 6G, the scarcity of spectrum resources is becoming increasingly prominent. Traditional static or semi-static resource allocation methods are no longer sufficient to meet the ever-increasing network capacity and diversified service demands. Static allocation typically relies on fixed spectrum allocation and preset resource scheduling strategies. While simple to implement, its utilization efficiency significantly decreases in scenarios with scarce spectrum resources or dynamically changing network loads. To address this, cognitive radio technology has emerged, enabling real-time monitoring of the spectrum environment and adaptive resource allocation through dynamic spectrum sensing. Dynamic spectrum sensing allows devices to dynamically select available frequency bands based on current spectrum usage, thereby improving spectrum utilization. In recent years, with the integration of artificial intelligence and big data analytics, resource allocation methods have gradually moved towards intelligence, such as machine learning-based spectrum prediction and graph theory-based network optimization techniques. However, the adaptability and optimization capabilities of these technologies in complex dynamic environments still need further improvement.

[0003] Therefore, although dynamic spectrum sensing has made some progress in the field of resource allocation, existing technologies still have many shortcomings in practical applications. First, traditional dynamic spectrum sensing methods are mostly based on static or low-dynamic assumptions, making it difficult to effectively capture the spatiotemporal fluctuations caused by high-speed device movement and rapid changes in network topology, thus limiting the real-time performance and accuracy of resource allocation strategies. Second, in complex network environments, interference relationships between devices often exhibit high-order nonlinear characteristics, while existing methods typically use low-order approximations, ignoring the deep structure of multidimensional interference and limiting the optimization depth of resource allocation. Furthermore, traditional optimization algorithms suffer from high computational complexity and are prone to getting trapped in local optima when facing large-scale networks and multidimensional resource allocation problems, making it difficult to meet the strict constraints of quality of service requirements. Summary of the Invention

[0004] The purpose of this section is to outline some aspects of embodiments of the present invention and to briefly describe some preferred embodiments. Simplifications or omissions may be made in this section, as well as in the abstract and title of this application, to avoid obscuring the purpose of these documents; however, such simplifications or omissions should not be construed as limiting the scope of the invention.

[0005] In view of the aforementioned existing problems, this invention is proposed. Therefore, this invention provides a wireless communication resource allocation method based on dynamic spectrum sensing to solve the problems mentioned in the background art.

[0006] To solve the above-mentioned technical problems, the present invention provides the following technical solution: a wireless communication resource allocation method based on dynamic spectrum sensing, comprising:

[0007] Acquire device node data and network environment parameters, preprocess the device node data and network environment parameters, and generate preprocessed feature data;

[0008] Based on the preprocessed feature data and time window, a spatiotemporal feature map is generated using dynamic spatiotemporal feature encoding.

[0009] Using device location information and hypergraph generation parameters, a hyperedge set is constructed through a dynamic hyperedge generation mechanism;

[0010] Based on the hyperedge set, transmit power, and channel gain, a high-order interference tensor is constructed and decomposed.

[0011] Based on the interference tensor, the number of resource blocks, and the quality of service requirements, spectral clustering and optimization algorithms are used to generate the optimal resource allocation strategy.

[0012] As a preferred embodiment of the wireless communication resource allocation method based on dynamic spectrum sensing described in this invention, the device node data includes location information, moving speed vector, transmit power, channel gain, and spectrum occupancy mode.

[0013] As a preferred embodiment of the wireless communication resource allocation method based on dynamic spectrum sensing described in this invention, the network environment parameters include noise density, bandwidth, and time window.

[0014] As a preferred embodiment of the wireless communication resource allocation method based on dynamic spectrum sensing described in this invention, the preprocessing includes normalizing and structuring the device node data and network environment parameters.

[0015] As a preferred embodiment of the wireless communication resource allocation method based on dynamic spectrum sensing described in this invention, the dynamic spatiotemporal feature encoding employs a spatiotemporal attention mechanism to calculate feature weights.

[0016] As a preferred embodiment of the wireless communication resource allocation method based on dynamic spectrum sensing described in this invention, the hyperedge dynamic generation mechanism is based on Delaunay partitioning and Transform the calculation of geodetic distance.

[0017] As a preferred embodiment of the wireless communication resource allocation method based on dynamic spectrum sensing described in this invention, the higher-order interference tensor is subjected to dimensionality reduction processing through Tucker decomposition.

[0018] As a preferred embodiment of the wireless communication resource allocation method based on dynamic spectrum sensing described in this invention, the optimization algorithm is a quantum approximation optimization algorithm.

[0019] As a preferred embodiment of the wireless communication resource allocation method based on dynamic spectrum awareness described in this invention, the quality of service requirements include latency and throughput requirements.

[0020] Compared with existing technologies, the beneficial effects of the invention are as follows:

[0021] 1. This invention, through dynamic spatiotemporal feature encoding and hyperedge dynamic generation mechanism, can capture spatiotemporal feature fluctuations caused by high-speed device movement and rapid changes in network topology in real time. Compared with the shortcomings of traditional dynamic spectrum sensing methods that are difficult to adapt to dynamic environments, it ensures the real-time performance and accuracy of resource allocation strategies, thereby maintaining high efficiency in complex and ever-changing network scenarios.

[0022] 2. By adopting high-order interference tensor decomposition technology, it is possible to effectively analyze and extract complex high-order interference relationships between devices and generate core interference patterns. This overcomes the limitations of existing technologies in high-order interference modeling, improves the optimization depth of resource allocation, and lays the foundation for more refined resource management.

[0023] 3. By combining spectral clustering and optimization algorithms, this invention generates a globally optimal resource allocation strategy for large-scale networks and multi-dimensional resource allocation problems. Compared with traditional optimization algorithms, this invention can reduce computational complexity and avoid the strategy from getting stuck in local optima, thereby meeting the service quality requirements of low latency and high throughput. Attached Figure Description

[0024] To more clearly illustrate the technical solutions of the embodiments of the present invention, the accompanying drawings used in the description of the embodiments will be briefly introduced below. Obviously, the drawings described below are only some embodiments of the present invention. For those skilled in the art, other drawings can be obtained based on these drawings without creative effort. Wherein:

[0025] Figure 1 This is a flowchart illustrating the overall process of a wireless communication resource allocation method based on dynamic spectrum sensing, according to an embodiment of the present invention. Detailed Implementation

[0026] To make the above-mentioned objects, features, and advantages of the present invention more apparent and understandable, specific embodiments of the present invention will be described in detail below with reference to the accompanying drawings. Obviously, the described embodiments are only a part of the embodiments of the present invention, and not all of them. Based on the embodiments of the present invention, all other embodiments obtained by those skilled in the art without creative effort should fall within the protection scope of the present invention.

[0027] Many specific details are set forth in the following description in order to provide a full understanding of the invention. However, the invention may also be practiced in other ways different from those described herein, and those skilled in the art can make similar extensions without departing from the spirit of the invention. Therefore, the invention is not limited to the specific embodiments disclosed below.

[0028] Secondly, the term "one embodiment" or "embodiment" as used herein refers to a specific feature, structure, or characteristic that may be included in at least one implementation of the present invention. The phrase "in one embodiment" appearing in different places in this specification does not necessarily refer to the same embodiment, nor is it a single or selective embodiment that is mutually exclusive with other embodiments.

[0029] This invention is described in detail with reference to the schematic diagrams. When detailing the embodiments of this invention, for ease of explanation, the cross-sectional views illustrating the device structure may be partially enlarged, not adhering to the usual scale. Furthermore, the schematic diagrams are merely examples and should not be construed as limiting the scope of protection of this invention. In actual fabrication, the three-dimensional spatial dimensions of length, width, and depth should be included.

[0030] Furthermore, in the description of this invention, it should be noted that the terms "upper," "lower," "inner," and "outer," etc., indicate the orientation or positional relationship based on the orientation or positional relationship shown in the accompanying drawings. These terms are used solely for the convenience of describing the invention and for simplifying the description, and do not indicate or imply that the device or element referred to must have a specific orientation, or be constructed and operated in a specific orientation. Therefore, they should not be construed as limitations on the invention. In addition, the terms "first," "second," or "third" are used for descriptive purposes only and should not be construed as indicating or implying relative importance.

[0031] Unless otherwise explicitly specified and limited, the terms "installation," "connection," and "joining" in this invention should be interpreted broadly. For example, they can refer to fixed connections, detachable connections, or integral connections; similarly, they can refer to mechanical connections, electrical connections, or direct connections, or indirect connections through an intermediate medium, or internal connections between two components. Those skilled in the art can understand the specific meaning of the above terms in this invention based on the specific circumstances.

[0032] Example 1

[0033] Reference Figure 1This is the first embodiment of the present invention, which provides a method for allocating wireless communication resources based on dynamic spectrum awareness, including:

[0034] S1. Obtain device node data and network environment parameters, preprocess the device node data and network environment parameters, and generate preprocessed feature data.

[0035] Furthermore, the device node data obtained from the communication network includes location information, movement speed vector, transmit power, channel gain, and spectrum occupancy pattern;

[0036] Specifically, the location information includes the coordinates (x, y) of each device node i in three-dimensional space. i ,y i ,z i ), used to analyze the spatial relationships between devices;

[0037] Specifically, the movement velocity vector contains the velocity components of a single device in its x and y directions, reflecting the movement characteristics of the single device;

[0038] Specifically, the transmission power is expressed as the transmission power P of the device, which is affected by the coverage and interference of the signal.

[0039] Specifically, channel gain G ij Let G be the channel gain from device node i to device node j, which reflects the attenuation process of the signal during propagation between devices;

[0040] Specifically, spectrum occupancy patterns are represented as the spectrum resources currently used by multiple devices, expressed in binary encoding or vector form;

[0041] Furthermore, network environment parameters obtained from the communication network include noise density, bandwidth, and time window;

[0042] Specifically, noise density N0 includes the background noise level, which can affect the quality of signal transmission;

[0043] Specifically, bandwidth B represents the total bandwidth or subcarrier bandwidth of the communication system, which determines the capacity of communication network resource allocation;

[0044] Specifically, the time window represents the time span of feature encoding, set to a 320ms period for the 5G frame structure;

[0045] It should be explained that, since the features in the raw data usually have different units and ranges of values, for example, location coordinates are in meters or kilometers, while power is in watts, preprocessing is required to scale the raw data to a uniform range of values ​​in order to eliminate the differences in units.

[0046] Furthermore, the device node data and network environment parameters are normalized to generate preprocessed feature data;

[0047] Specifically, the preprocessing method is selected based on the above data, that is, according to the data type, one of the following is selected for processing: min-max normalization, Z-Score normalization, decimal scaling normalization, and logarithmic normalization.

[0048] Specifically, for selection operations, if the data follows a normal distribution, Z-Score normalization is used; if the data follows a skewed distribution, log normalization is used; if outliers are present, decimal scaling normalization is used; and if the data contains explicit upper and lower bounds, min-max normalization is used.

[0049] It should be noted that by using selective preprocessing, the performance of each data point in constructing dynamic spatiotemporal feature codes can be maximized;

[0050] It should be noted that because the normalized data is scattered and has inconsistent formats, such as tables, lists, or time series, directly using it to construct spatiotemporal feature codes would affect the construction efficiency. Therefore, it is necessary to organize this data in a structured manner. This structured organization can be broken down into structured format and data organization. Common structured formats include two-dimensional matrices, three-dimensional tensors, and multidimensional tensors. Data organization includes constructing feature matrices, organizing time series data, and organizing relational data. Processing the normalized data through structured organization can improve the efficiency of constructing dynamic spatiotemporal feature codes.

[0051] S2. Based on the preprocessed feature data and time window, a spatiotemporal feature map is generated using dynamic spatiotemporal feature encoding;

[0052] Furthermore, a three-dimensional feature tensor is constructed based on the preprocessed feature data and the time window in the network parameters. Where N is the number of device nodes, representing the total number of terminal devices in the communication network; T represents the time window; and 5 represents the five dimensions of features, namely instantaneous signal-to-interference-plus-noise ratio (SINR), mobile speed vector, spectrum occupancy pattern, transmit power, and quality of service (QoS) requirements.

[0053] Furthermore, for the signal received by device node j from device node i, its instantaneous signal-to-dryness ratio (SINR) ij It is obtained through the following formula:

[0054]

[0055] Among them, P i ×G ij The received power of the target signal is expressed as ∑k≠i P k ×G kj N0 represents the sum of interference power from all other device nodes to device node j, and N0×B represents the background noise power.

[0056] It should be noted that the interference power needs to be calculated by summing the transmit power and channel gain of all devices in the communication network, which depends on the integrity of the device node data.

[0057] It should be explained that the Spatio-Temporal Attention Mechanism is an extension of the attention mechanism, designed to capture important features of data in both spatial and temporal dimensions. In wireless communication networks, features such as device location, speed, and channel state change dynamically with time and space. Therefore, the Spatio-Temporal Attention Mechanism is needed to identify spatial correlations (such as interference from neighboring devices) and temporal continuity (such as trends in device movement trajectories) between devices.

[0058] Furthermore, a three-dimensional feature tensor will be constructed. The input is fed into the spatiotemporal attention mechanism, and the output is a spatiotemporal feature map;

[0059] Specifically, the spatiotemporal attention mechanism transforms the input three-dimensional feature tensor X through a linear transformation to generate a query matrix Q, a key matrix K, and a value matrix V. It then uses the query matrix Q and the key matrix K to calculate the attention score, processes it through the Softmax function to obtain the attention weight matrix, and finally sums the value matrix V according to the attention weights in the attention weight matrix to obtain the spatiotemporal feature map.

[0060] Specifically, the spatiotemporal attention mechanism α i,j This can be expressed as a mathematical formula:

[0061]

[0062] Where Q = X i W Q W Q The query weight matrix is ​​used to linearly transform the input feature tensor X into a query matrix Q, where K = X. j W K K is the key weight matrix, used to linearly transform the input feature tensor X into the key matrix K;

[0063] Specifically, W Q and W K The dimensions are all Where, d in The dimension of the input three-dimensional feature tensor is 5 in the scheme of this invention, d.k For the internal dimensions of the spatiotemporal attention mechanism;

[0064] It should be noted that, due to W Q and W K The generated Q and K are both used to calculate the attention score QK. T Therefore, the attention score is based on the dot product of Q and K, reflecting the degree of matching between query features and key features, while W... Q and W K This can determine the degree of matching; for example, if W Q Emphasizing SINR, while W K If the transmit power is emphasized, the corresponding spatiotemporal attention mechanism will pay more attention to the relationship between signal-to-noise ratio and power;

[0065] S3. Using device location information and hypergraph generation parameters, construct a hyperedge set through a dynamic hyperedge generation mechanism;

[0066] Furthermore, based on the device location information, a location code p is assigned to the device node in three-dimensional Euclidean space. i =(x i ,y i ,z i ,Δt), where Δt represents the time offset;

[0067] It should be noted that assigning location codes to device nodes is used to capture the device's position in three-dimensional Euclidean space;

[0068] Furthermore, adopt The transformation calculates the geodesic distance between device nodes in three-dimensional Euclidean space;

[0069] It needs to be explained that the distribution of devices and signal propagation in wireless communication networks typically do not perfectly conform to Euclidean geometry, and the locations of devices change dynamically over time; for example, communication signal attenuation often varies non-linearly with distance, or is subject to complex influences from environmental factors (such as obstacles); and A transformation is a conformal mapping on the complex plane, which is often used in hyperbolic geometry to keep angles and geodesic distances unchanged. The transform can provide a distance metric that more closely reflects the actual signal propagation characteristics through hyperbolic geometry modeling. Compared with the traditional Euclidean distance, it can amplify the signal differences between near-field devices while compressing the impact of far-field devices on signal propagation, which is consistent with the characteristics of near-field interference in wireless communication.

[0070] Specifically, the formula for calculating geodetic distance is as follows:

[0071]

[0072] Where, d M (i,j) represents the geodesic distance between device node i and device node j; |·| represents the Euclidean distance. 2 Represented as the square of the Euclidean modulus;

[0073] Furthermore, in three-dimensional Euclidean space, the three-dimensional coordinates of the input device are processed based on the three-dimensional Delaunay subdivision to generate a tetrahedron.

[0074] It should be explained that, since the devices in a wireless communication network are discretely distributed, the hyperedge needs to reflect the proximity relationship between devices; while Delaunay partitioning can automatically generate the optimal triangles or tetrahedrons through geometric constraints to ensure that adjacent devices are connected.

[0075] Furthermore, a threshold is set, and the density of the tetrahedron is calculated. When the density exceeds the threshold, a hyperedge containing the vertices of the tetrahedron is generated. The hyperedge weight of the tetrahedron vertices is calculated using the geodesic distance obtained above, and the hyperedge attenuation factor is obtained.

[0076] It should be noted that combining each generated hyperedge together forms the hyperedge set E;

[0077] Specifically, the density ρ of a tetrahedron is calculated as follows:

[0078]

[0079] Where V is the volume of the tetrahedron.

[0080] Specifically, set a threshold ρ thres Based on the load N of the communication network active Dynamically adjust equipment density:

[0081]

[0082] Where ρ0 represents the initial density of the tetrahedron, and N active N represents the number of active devices. max This represents the maximum capacity of the resource block.

[0083] Specifically, the super-edge attenuation factor β a Represented as:

[0084]

[0085] Where λ is the learning rate parameter, Var(d M (i,j)) represents the variance of the geodesic distance, a represents the hyperedge, and e represents the fundamental constant of the natural logarithm;

[0086] It should be noted that, since the generation of hyperedges requires topological relationships based on distance, through... Transformation can solve this problem well, and the use of 3D Delaunay partitioning can update the tetrahedron according to the real-time changes in device coordinates. That is, when the device moves, the partitioning results can be automatically adjusted, and the hyperedge set E is also updated accordingly, which meets the requirements of dynamic networks.

[0087] S4. Based on the hyperedge set, transmit power, and channel gain, construct a high-order interference tensor and decompose it;

[0088] Furthermore, based on the hyperedge set E, transmit power P, and channel gain G obtained above, a higher-order interference tensor of the hypergraph is constructed.

[0089]

[0090] Where each element A of tensor A i,j,k All represent the interference intensity between device nodes i, j, k; while I1×I2×I3×…×I N This indicates that there are several device nodes in the communication network, which are represented by a high-order tensor and a ternary interference relationship.

[0091] It should be explained that, due to the large amount of data in the constructed high-order disturbance tensor A, direct manipulation would increase computational complexity. Therefore, it is necessary to decompose it and extract the core features to reduce data redundancy.

[0092] Furthermore, the higher-order disturbance tensor A is decomposed using the Tucker decomposition method, yielding:

[0093] A≈H×1U (1) ×2U (2) ×3U (3) ×…× n U (n)

[0094] in, Represented as a core tensor, its dimension is smaller than the original higher-order interference tensor A, and it is used to extract the main patterns of interference between devices. Represented as an n-dimensional factor matrix, it describes the characteristic orientation of each device pair; R n It is the rank after dimensional reduction, × n This can be represented as tensor-matrix multiplication along the nth dimension;

[0095] It should be noted that in wireless communication, interference may be concentrated on certain device pairs or superedges, and by adopting Tucker decomposition, the main patterns of these interferences can be preserved, while compressing the irrelevant dimensions of device nodes.

[0096] S5. Based on the interference tensor, the number of resource blocks, and the quality of service requirements, spectral clustering and optimization algorithms are used to generate the optimal resource allocation strategy.

[0097] It needs to be explained that a hyperedge may represent multi-party interference caused by a group of devices sharing the spectrum. Since the hypergraph is composed of hyperedges, it is necessary to capture the topology and interference characteristics between devices in the communication network in the hypergraph.

[0098] Furthermore, the Laplacian matrix L is constructed, and its calculation method for the hypergraph is as follows:

[0099] L=DA (m)

[0100] Where D represents the degree matrix, and its diagonal elements represent the degree of each device node (i.e., the number of connected hyperedges e); A (m) It is represented as the disturbance tensor reduced to order m after decomposition;

[0101] Furthermore, by performing eigenvalue decomposition on the Laplace matrix L and solving the generalized characteristic equation, we obtain:

[0102] Lv=μDv

[0103] Among them, the first F smallest non-zero eigenvalues ​​(μ1, μ2, ..., μ) are selected. F ) and their corresponding feature vectors (v1, v2, ..., v F ), where F is the preset number of clusters;

[0104] It should be noted that the extracted non-zero eigenvalues ​​and their corresponding eigenvectors reflect the connectivity of the hypergraph, and the eigenvector corresponding to the smallest non-zero eigenvalue shows the direction of natural grouping;

[0105] Furthermore, the F smallest eigenvectors are stacked into an N×F matrix M, where each row of matrix M represents the coordinates of a device node in the F-dimensional embedding space, so that devices with similar interference characteristics are in close proximity in the embedding space.

[0106] Specifically, in the embedded space, the device nodes are divided into F clusters using a traditional clustering algorithm. The device nodes in each cluster are allocated the same resource blocks because they have similar interference characteristics.

[0107] Furthermore, based on the resource block allocation results, considering the minimization of device interference and the maximization of resource utilization efficiency, an objective function is established:

[0108]

[0109] Wherein, the resource block allocation vector, S i ω represents the resource block allocated to device node i.a The weight of hyperedge a is the interference intensity; δ(a,S) represents the hyperedge cutting function. If the resource blocks allocated to the devices within hyperedge a are different, then δ(a,S) = 1, otherwise δ(a,S) = 0.

[0110] Specifically, the optimization problem of the above objective function is solved by using the Quantum Approximation Optimization Algorithm (QAOA);

[0111] It should be noted that, because QAOA itself utilizes the advantages of quantum computing, it can find the global optimal or near-optimal solution in a complex search space, thus breaking through the local optimal limitation of traditional greedy algorithms.

[0112] Furthermore, the results obtained by the quantum approximation optimization algorithm are passed to the MAC layer scheduler to guide the specific resource block allocation and execution.

[0113] Those skilled in the art will understand that embodiments of the present invention can be provided as methods, systems, or computer program products. Therefore, this application can take the form of a completely hardware embodiment, a completely software embodiment, or an embodiment combining software and hardware aspects. Furthermore, this application can take the form of a computer program product implemented on one or more computer-usable storage media (including but not limited to disk storage, CD-ROM, optical storage, etc.) containing computer-usable program code. The solutions in the embodiments of this application can be implemented using various computer languages, such as the object-oriented programming language Java and the interpreted scripting language JavaScript.

[0114] This application is described with reference to flowchart illustrations and / or block diagrams of methods, apparatus (systems), and computer program products according to embodiments of this application. It will be understood that each block of the flowchart illustrations and / or block diagrams, and combinations of blocks in the flowchart illustrations and / or block diagrams, can be implemented by computer program instructions. These computer program instructions can be provided to a processor of a general-purpose computer, special-purpose computer, embedded processor, or other programmable data processing apparatus to produce a machine, such that the instructions, which execute via the processor of the computer or other programmable data processing apparatus, generate instructions for implementing the flowchart... Figure 1 One or more processes and / or boxes Figure 1 A device that provides the functions specified in one or more boxes.

[0115] These computer program instructions may also be stored in a computer-readable storage medium that can direct a computer or other programmable data processing device to function in a particular manner, such that the instructions stored in the computer-readable storage medium produce an article of manufacture including instruction means, which are implemented in a process Figure 1 One or more processes and / or boxes Figure 1The function specified in one or more boxes.

[0116] These computer program instructions may also be loaded onto a computer or other programmable data processing equipment to cause a series of operational steps to be performed on the computer or other programmable equipment to produce a computer-implemented process, thereby providing instructions that execute on the computer or other programmable equipment for implementing the process. Figure 1 One or more processes and / or boxes Figure 1 The steps of the function specified in one or more boxes.

[0117] Although preferred embodiments of this application have been described, those skilled in the art, upon learning the basic inventive concept, can make other changes and modifications to these embodiments. Therefore, the appended claims are intended to be interpreted as including the preferred embodiments as well as all changes and modifications falling within the scope of this application.

[0118] Obviously, those skilled in the art can make various modifications and variations to this application without departing from the spirit and scope of this application. Therefore, if such modifications and variations fall within the scope of the claims of this application and their equivalents, this application also intends to include such modifications and variations.

Claims

1. A method for dynamic spectrum sensing based wireless communication resource allocation, characterized in that, The method comprises the following steps: Obtaining device node data and network environment parameters, preprocessing the device node data and network environment parameters, and generating preprocessed feature data; Based on the preprocessed feature data and the time window, a dynamic spatio-temporal feature encoding is used to generate a spatio-temporal feature map; Using device location information and hypergraph generation parameters, a hyperedge set is constructed through a hyperedge dynamic generation mechanism; Based on the hyperedge set, the transmit power and the channel gain, a high-order interference tensor is constructed and decomposed; According to the interference tensor, the number of resource blocks and the quality of service requirements, a spectral clustering and optimization algorithm is used to generate an optimal resource allocation strategy.

2. The dynamic spectrum sensing based radio communication resource allocation method of claim 1, wherein, The device node data includes location information, moving speed vector, transmit power, channel gain and spectrum occupation mode.

3. The dynamic spectrum sensing based radio communication resource allocation method of claim 1, wherein, The network environment parameters include noise density, bandwidth and time window.

4. The dynamic spectrum sensing based radio communication resource allocation method of claim 1, wherein, The preprocessing includes normalization processing and structured organization of the device node data and network environment parameters.

5. The dynamic spectrum sensing based radio communication resource allocation method of claim 1, wherein, The dynamic spatio-temporal feature encoding uses a spatio-temporal attention mechanism to calculate feature weights.

6. The dynamic spectrum sensing based radio communication resource allocation method of claim 1, wherein, The super-edge dynamic generation mechanism is based on Delaunay partitioning and Transform calculates geodesic distance.

7. The dynamic spectrum sensing based radio communication resource allocation method of claim 1, wherein, The high-order interference tensor is processed by Tucker decomposition for dimension reduction.

8. The dynamic spectrum sensing based radio communication resource allocation method of claim 1, wherein, The optimization algorithm is a quantum approximate optimization algorithm.

9. The dynamic spectrum sensing based radio communication resource allocation method of claim 1, wherein, The quality of service requirements include delay and throughput requirements.

Images