Spectrum map communication organization method, device and equipment based on semantic information

Abstract

The present application relates to a kind of method, device and equipment based on semantic information spectrum map communication organization.The method includes: obtaining spectrum map and text information;Using variational autoencoder to encode the spectrum map, obtain image feature information, image feature information is mapped to latent feature space, and image latent space vector is obtained;Text information is encoded, and text feature is obtained;According to the attribute of text prompt word in the same batch feature dimension, text prompt word feature vector is obtained by stacking;Based on denoising diffusion implicit model, image latent space vector is added noise, and the noise image feature of adding noise is obtained;After the noise image feature and text prompt word feature vector are fused, they are input into UNet network for noise prediction together with time step number, and image latent space vector is optimized according to the predicted noise data, to obtain real spectrum map.The recombination energy efficiency of spectrum map can be improved by using the method.

Description

TECHNICAL FIELD

[0001] The present application relates to the technical field of wireless communication networks, in particular to a spectrum map communication organization method, device and equipment based on semantic information. BACKGROUND

[0002] Under the background of rapid development of 6G, intelligent networking and smart city, a large number of wireless communication devices are widely deployed, which leads to rapid complexification of electromagnetic spectrum environment and increasingly sharp contradiction of frequency use. Electromagnetic spectrum, as a key carrier and valuable resource of network information transmission, has important significance for monitoring and allocating spectrum resources, avoiding channel conflict and interference, optimizing transmission power, frequency selection and signal coding strategy to improve communication efficiency and reduce energy consumption.

[0003] Spectrum map is a key tool for understanding complex electromagnetic environment, which describes electromagnetic environment from multiple dimensions such as time, frequency, space and power by combining geographic information system, supports effective acquisition of electromagnetic spectrum resource distribution, and mining of multi-dimensional reusable value of spectrum. Spectrum map assists wireless resource management in an intuitive and flexible way, and has been widely used in the fields of cognitive radio, interference management and proactive resource allocation in expected network. However, constructing spectrum map is a complex computing task, and the distribution of spectrum map from cloud server to user needs to consume a lot of bandwidth and time, and in the rapidly changing electromagnetic environment, such large bandwidth consumption may bring a heavy burden to the communication system. SUMMARY

[0004] Therefore, it is necessary to provide a spectrum map communication organization method, device and equipment based on semantic information, which can improve the energy efficiency of spectrum map communication organization.

[0005] A spectrum map communication organization method based on semantic information, the method is implemented in the mutual communication between cloud / edge server and frequency user, comprising:

[0006] Obtain spectrum map and text information.

[0007] Encode the spectrum map using variational autoencoder to obtain image feature information, map the image feature information to latent feature space to obtain image latent space vector.

[0008] Encode the text information to obtain text features, stack the text prompt word features in the same batch of feature dimensions according to the attributes of the text prompt words to obtain text prompt word feature vector.

[0009] Based on the denoising diffusion implicit model, the image latent space vector is added with noise to obtain the noise-added image feature.

[0010] The noise-added image feature and the text prompt feature vector are fused, and are input into a UNet network together with the time step number for noise prediction, the image latent space vector is optimized according to the predicted noise data, and a real spectrum map is obtained.

[0011] In one of the embodiments, the text information is encoded by using CLIP text vectorization of a CLIP text encoder to obtain encoded text information, and the encoded text information is input into a CLIP text model for feature extraction to obtain text features.

[0012] In one of the embodiments, the text information is encoded by using CLIP text vectorization of a CLIP text encoder to obtain encoded text information, and the encoded text information is input into a CLIP text model for feature extraction to obtain text features.

[0013] In one of the embodiments, the text information is encoded by using CLIP text vectorization of a CLIP text encoder to obtain encoded text information, and the encoded text information is input into a CLIP text model for feature extraction to obtain text features.

[0014] In one of the embodiments, the text information is encoded by using CLIP text vectorization of a CLIP text encoder to obtain encoded text information, and the encoded text information is input into a CLIP text model for feature extraction to obtain text features.

[0015] In one of the embodiments, the text information is encoded by using CLIP text vectorization of a CLIP text encoder to obtain encoded text information, and the encoded text information is input into a CLIP text model for feature extraction to obtain text features.

[0016] In one of the embodiments, the text information is encoded by using CLIP text vectorization of a CLIP text encoder to obtain encoded text information, and the encoded text information is input into a CLIP text model for feature extraction to obtain text features.

[0017] A spectrum map communication organization device based on semantic information is loaded in a mutual communication system implemented in a cloud / edge server and a frequency user, and the device comprises:

[0018] An information acquisition module is configured to acquire a spectrum map and text information.

[0019] An image latent space vector acquisition module is configured to encode the spectral map by using a variational autoencoder, obtain image feature information, map the image feature information to a latent feature space, and obtain an image latent space vector.

[0020] A text prompt word feature vector acquisition module is configured to encode the text information, obtain text features, stack the text features according to the attributes of the text prompt words in the same batch of feature dimensions, and obtain a text prompt word feature vector.

[0021] A noisy image feature acquisition module is configured to add noise to the image latent space vector based on a denoising diffusion implicit model, and obtain a noisy image feature.

[0022] A real spectral map construction module is configured to fuse the noisy image feature and the text prompt word feature vector, input the fusion result and a time step number to a UNet network for noise prediction, optimize the image latent space vector according to the predicted noise data, and obtain a real spectral map.

[0023] A computer device includes a memory and a processor, the memory stores a computer program, and the processor implements the following steps when executing the computer program:

[0024] A spectral map and text information are obtained.

[0025] The spectral map is encoded by using a variational autoencoder, image feature information is obtained, the image feature information is mapped to a latent feature space, and an image latent space vector is obtained.

[0026] The text information is encoded, text features are obtained, and the text features are stacked according to the attributes of the text prompt words in the same batch of feature dimensions, and a text prompt word feature vector is obtained.

[0027] The image latent space vector is added with noise based on a denoising diffusion implicit model, and a noisy image feature is obtained.

[0028] The noisy image feature and the text prompt word feature vector are fused, and the fusion result and a time step number are input to a UNet network for noise prediction, the image latent space vector is optimized according to the predicted noise data, and a real spectral map is obtained.

[0029] The above method, device and equipment for organizing spectrum map communication based on semantic information first encodes the spectrum map through VAE, extracts image features and maps them to a latent feature space to generate an image latent space vector. This process not only effectively compresses the image information, but also extracts the most representative features, making subsequent processing more efficient. At the same time, the text information is encoded to extract its features, and stacked on the feature dimension of the same batch according to the attributes of the text prompt word to form a text prompt word feature vector. This step ensures accurate expression of text information in the feature space, allowing it to seamlessly integrate with image features. Next, the image latent space vector is subjected to noise processing based on the denoising diffusion implicit model. Noise processing simulates noise interference in the actual environment, allowing the model to better adapt to and handle noise conditions in the real world. In this way, the model can learn how to extract useful signals from noise, improving its robustness and generalization ability. After the noisy image features are fused with the text prompt word feature vector, they are input into the UNet network along with the time step number for noise prediction. UNet network is a commonly used deep learning architecture that excels at image segmentation and reconstruction tasks. In this technical solution, the UNet network is used to predict noise data and optimize the image latent space vector based on the prediction results. Through multiple iterations and optimization, a real and high-quality spectrum map can be generated. BRIEF DESCRIPTION OF DRAWINGS

[0030] Figure 1 Flowchart of a method for organizing spectrum map communication based on semantic information in an embodiment;

[0031] Figure 2 Schematic diagram of a processing framework for spectrum map communication based on semantic information in an embodiment;

[0032] Figure 3 Flowchart of image feature processing in a spectrum map based on semantic information in an embodiment;

[0033] Figure 4 Flowchart of text feature processing in a spectrum map based on semantic information in an embodiment;

[0034] Figure 5 Block diagram of a device for organizing spectrum map communication based on semantic information in an embodiment;

[0035] Figure 6 Internal structure diagram of a computer device in an embodiment. DETAILED DESCRIPTION

[0036] In order to make the purposes, technical solutions and advantages of the present application clearer, the present application is further described in detail below in combination with the drawings and examples. It should be understood that the specific examples described herein are only used to explain the present application and do not limit the present application.

[0037] In one embodiment, as shown in Figure 1 A method for organizing spectrum map communication based on semantic information is provided, which is implemented in the mutual communication of a cloud / edge server and a frequency user, and includes the following steps:

[0038] Step 102: Obtain a spectrum map and text information.

[0039] Step 104: Encode the spectrum map using a variational autoencoder to obtain image feature information, map the image feature information to a latent feature space, and obtain an image latent space vector.

[0040] Step 106: Encode the text information to obtain text features, stack the text prompt word features in the same batch of feature dimensions according to the attributes of the text prompt words, and obtain a text prompt word feature vector.

[0041] Step 108: Add noise to the image latent space vector based on a denoising diffusion implicit model to obtain a noisy image feature.

[0042] Step 110: Fuse the noisy image feature and the text prompt word feature vector, input them to a UNet network together with the time step number for noise prediction, optimize the image latent space vector according to the predicted noise data, and obtain a real spectrum map.

[0043] In the above method of organizing spectrum map communication based on semantic information, first, the spectrum map is encoded by VAE to extract image features and map them into a latent feature space, generating an image latent space vector. This process not only effectively compresses the image information, but also extracts the most representative features, making subsequent processing more efficient. At the same time, the text information is encoded to extract its features, and stacked in the same batch of feature dimensions according to the attributes of the text prompt words, forming a text prompt word feature vector. This step ensures accurate expression of text information in the feature space, allowing it to seamlessly integrate with image features. Next, the image latent space vector is subjected to noise processing based on the denoising diffusion implicit model. Noise processing simulates noise interference in the actual environment, enabling the model to better adapt to and handle noise in the real world. In this way, the model can learn how to extract useful signals from noise, improving its robustness and generalization ability. After the noisy image features are fused with the text prompt word feature vector, they are input into the UNet network along with the time step number for noise prediction. UNet network is a commonly used deep learning architecture that excels at image segmentation and reconstruction tasks. In this technical solution, the UNet network is used to predict noise data and optimize the image latent space vector based on the prediction results. Through multiple iterations and optimizations, a real and high-quality spectrum map can be generated.

[0044] In one embodiment, as shown in Figure 2 A processing block diagram for organizing spectrum map communication based on semantic information is provided, which is implemented in the mutual communication between cloud / edge servers and users, and includes the following steps:

[0045] The text information is encoded by the CLIP text encoder, and the original spectrum picture is encoded by the variational autoencoder and then added with noise in the scheduler of the diffusion model. The obtained text features and the semantic picture after adding noise to the generated picture features are transmitted to the user through the channel. The user combines the time step number and the text embedding to calculate the current time noise, predicts the noise, removes the noise, and restores the original high-definition spectrum map. Based on the DDIM diffusion model, the noise value can be understood as the proportion of noise that needs to be added, i.e., the proportion of the picture that needs to be reconstructed. It can be considered as the reconstruction noise value percentage, and the remaining (1-noise value) percentage features are retained. By setting the noise value and the total number of steps, the input picture will be added with noise for noise value x total number of steps, and the remaining steps will not be added with noise. If only the text information or the text information and the picture semantic information are used to construct the spectrum map using the unconditional limited diffusion model, it is meaningless for the precise spectrum map task construction. Only the conditional limited diffusion model can be used for precise spectrum map construction. Compared with the transmission of original spectrum map, the transmission of text information and picture semantic information effectively reduces the required data volume, and improves the recombination efficiency of the spectrum map.

[0046] In one embodiment, as shown in Figure 3 A text feature processing flow of a spectrum map based on semantic information is provided, which is implemented in mutual communication between a cloud / edge server and a frequency user, and includes the following steps:

[0047] Step S300, high-definition spectrum map and text information capable of representing spectrum map features, such as transmitter location, power, and other information, are obtained.

[0048] Step S310, a variational autoencoder is used for encoding to obtain feature information of the image, and the image feature information is mapped to a latent feature space.

[0049] Step S320, a contrastive language-image pre-training (CLIP) model is used for text encoding on the input text to obtain text features.

[0050] Step S330, after obtaining the high-definition spectrum map image features, noise is added based on a denoising diffusion implicit model (DDIM) to obtain features of the input picture after noise addition.

[0051] Step S340, after the obtained text features and the generated picture noise features are fused and input into a UNet network, the number of time steps and the text embedding are combined to calculate the current time noise, and the noise is predicted.

[0052] Step S150, based on the obtained latent space noise-added image, the latent space features are decoded using a variational autoencoder to generate an accurate high-definition spectrum picture.

[0053] In this embodiment, the communication between the cloud / edge server and the frequency user, the transmission of the spectrum map semantic information, and the user using a diffusion model to construct the spectrum map based on the text information and the semantic noise picture information are used. The conditional diffusion model uses text information and captured image semantic information to perform an accurate spectrum map reconstruction task. This method can reduce other redundant information during spectrum map construction and distribution, and can accurately construct the spectrum map under the condition of only transmitting semantic information, thereby significantly reducing the communication overhead of data transmission while ensuring accuracy.

[0054] In one embodiment, as shown in Figure 4 A text feature processing flow of a spectrum map based on semantic information is provided, which is implemented in mutual communication between a cloud / edge server and a frequency user, and includes the following steps:

[0055] In step S400, for the input text prompt word information including positive and negative prompt words, the input positive prompt word feature vector and negative prompt word feature vector are stacked in the batch feature dimension.

[0056] In step S410, after stacking, the obtained image latent space vector, step number and text prompt word feature vector are input into the UNet network, and the result is segmented in the batch feature dimension.

[0057] In step S420, for the negative prompt word, the picture to be constructed is expected to be away from its influence and biased towards the influence of the positive prompt word. To this end, we calculate the distance between the two to be expanded.

[0058] In step S430, for the UNet network, after predicting the noise at each step, new Gaussian noise is added to the network according to a certain proportion. By introducing new noise, the robustness of the network can be increased by iteratively predicting and denoising the noise.

[0059] In step S440, by sampling in the latent space multiple times, the required spectral map is obtained by decoding the obtained latent vector using a variational autoencoder.

[0060] In this embodiment, the user decodes the required spectral map according to the text information features and the image information features after adding noise.

[0061] In one embodiment, the text information is encoded using CLIP text vectorization of the CLIP text encoder to obtain encoded text information, and the encoded text information is input into the CLIP text model for feature extraction to obtain text features.

[0062] In one embodiment, the prompt words in the text information are classified according to a predetermined rule to obtain positive prompt words and negative prompt words, and the positive prompt words and the negative prompt words are stacked in the same batch feature dimension to obtain a text prompt word feature vector.

[0063] In one embodiment, a noise value and a total number of noise iteration steps are set, the image latent space vector is added with noise according to the product of the noise value and the total number of noise iteration steps, the difference between the total number of noise iteration steps and the product is calculated to obtain a remaining iteration step number, and the image latent space vector is not added with noise in the remaining iteration step number.

[0064] In one embodiment, a value with the same reconstruction ratio as the spectrum map is selected as the noise value, and the latent space vector of the image is denoised based on the denoising diffusion implicit model to reconstruct the noise value percentage and obtain the denoised image feature corresponding to the noise percentage feature.

[0065] In one embodiment, the noisy image features and the text prompt word feature vectors are fused using a neural network model, and then input into the UNet network along with the time steps obtained through the fully connected layer of the neural network for noise prediction.

[0066] In one embodiment, after each step of the UNet network predicts noise, new Gaussian noise is obtained. The Gaussian noise is input into the UNet network for denoising iteration according to a preset ratio. The image latent space vector is optimized by sampling in the latent space multiple times, and the real spectrum map is reconstructed based on the optimized image latent space vector.

[0067] It should be understood that, although Figure 1 , 3 The steps in the flowchart are shown sequentially as indicated by the arrows, but these steps are not necessarily executed in the order indicated by the arrows. Unless otherwise specified herein, there is no strict order in which these steps are executed, and they can be performed in other orders. Figure 1 , 3 At least some of the steps in the process may include multiple sub-steps or multiple stages. These sub-steps or stages are not necessarily completed at the same time, but can be executed at different times. The execution order of these sub-steps or stages is not necessarily sequential, but can be executed in turn or alternately with other steps or at least some of the sub-steps or stages of other steps.

[0068] In one embodiment, such as Figure 5 As shown, a spectrum map communication organization device based on semantic information is provided, including: an information acquisition module 502, an image latent space vector acquisition module 504, a text prompt word feature vector acquisition module 506, a noisy image feature acquisition module 508, and a real spectrum map construction module 510, wherein:

[0069] Information acquisition module 502 is used to acquire spectrum map and text information;

[0070] The image latent space vector acquisition module 504 is used to encode the spectrum map using a variational autoencoder to obtain image feature information, and to map the image feature information to the latent feature space to obtain the image latent space vector.

[0071] The text prompt word feature vector acquisition module 506 is configured to encode the text information to obtain text features, stack the attributes of the text prompt words in the same batch according to the text features in the feature dimension, and obtain a text prompt word feature vector.

[0072] The noisy image feature acquisition module 508 is configured to add noise to the image latent space vector based on the denoising diffusion implicit model to obtain a noisy image feature.

[0073] The real spectral map construction module 510 is configured to fuse the noisy image feature and the text prompt word feature vector, input them to a UNet network together with a time step number for noise prediction, and optimize the image latent space vector according to the predicted noise data to obtain a real spectral map.

[0074] The specific definitions of the device for organizing spectral map communication based on semantic information can refer to the definitions of the method for organizing spectral map communication based on semantic information, which will not be described here. The various modules in the device for organizing spectral map communication based on semantic information can be realized by software, hardware, and combinations thereof, in whole or in part. The various modules described above can be embedded in or independent of the processor in the computer device in hardware form, or can be stored in the memory in the computer device in software form, so as to be called and executed by the processor to perform the operations corresponding to the various modules.

[0075] In one embodiment, a computer device is provided, which can be a terminal, and an internal structure diagram thereof can be as shown in Figure 6 The computer device includes a processor, a memory, a network interface, a display screen, and an input device connected through a system bus. The processor of the computer device is configured to provide computing and control capabilities. The memory of the computer device includes a non-volatile storage medium and an internal memory. The non-volatile storage medium stores an operating system and a computer program. The internal memory provides an environment for the operating system and the computer program in the non-volatile storage medium to run. The network interface of the computer device is configured to communicate with external terminals through network connection. The computer program is executed by the processor to implement a method for organizing spectral map communication based on semantic information. The display screen of the computer device can be a liquid crystal display screen or an electronic ink display screen. The input device of the computer device can be a touch layer overlaid on the display screen, or can be a key, trackball, or touchpad arranged on the shell of the computer device, or can be an external keyboard, touchpad, or mouse, etc.

[0076] Those skilled in the art can understand that the structure shown in the figure Y is only a block diagram of part of the structure related to the scheme of the present application, and does not constitute a limitation on the computer device to which the scheme of the present application is applied. The specific computer device can include more or fewer components than those shown in the figure, or combine certain components, or have a different arrangement of components.

[0077] In one embodiment, a computer device is provided, comprising a memory and a processor, the memory storing a computer program, and the processor implementing the following steps when executing the computer program:

[0078] Obtaining a spectrum map and text information.

[0079] Encoding the spectrum map using a variational autoencoder to obtain image feature information, mapping the image feature information to a latent feature space to obtain an image latent space vector.

[0080] Encoding the text information to obtain text features, and stacking the attributes of the text prompt words in the same batch of feature dimensions according to the text features to obtain a text prompt word feature vector.

[0081] Based on a denoising diffusion implicit model, the image latent space vector is added with noise to obtain a noisy image feature.

[0082] After the noisy image feature and the text prompt word feature vector are fused, they are input into a UNet network together with the time step number for noise prediction, and the image latent space vector is optimized according to the predicted noise data to obtain a real spectrum map.

[0083] Those skilled in the art can understand that all or part of the processes in the above-mentioned embodiment methods can be completed by instructing the relevant hardware through a computer program. The computer program can be stored in a non-volatile computer readable storage medium, and when the computer program is executed, the processes of the above-mentioned embodiments of the methods can be included. Any reference to memory, storage, database or other medium used in the embodiments provided by the present application can include non-volatile and / or volatile memory. Non-volatile memory can include read-only memory (ROM), programmable ROM (PROM), electrically programmable ROM (EPROM), electrically erasable programmable ROM (EEPROM) or flash memory. Volatile memory can include random access memory (RAM) or external cache memory. As an illustration but not limitation, RAM is available in various forms, such as static RAM (SRAM), dynamic RAM (DRAM), synchronous DRAM (SDRAM), double data rate SDRAM (DDR SDRAM), enhanced SDRAM (ESDRAM), synchronous link (Synchlink) DRAM (SLDRAM), memory bus (Rambus) direct RAM (RDRAM), direct memory bus dynamic RAM (DRDRAM), and memory bus dynamic RAM (RDRAM), etc.

[0084] Any combination of the technical features of the above embodiments can be made. In order to make the description simple, all possible combinations of the technical features in the above embodiments are not described, however, as long as the combination of the technical features does not exist, it should be considered as the scope of the present application.

[0085] The above embodiments only express several embodiments of the present application, and the description is more specific and detailed, but it should not be understood as a limitation on the scope of the patent. It should be pointed out that for ordinary skilled in the art, without departing from the concept of the present application, a number of modifications and improvements can be made, which are within the scope of the present application. Therefore, the scope of protection of the present application should be subject to the appended claims.

Claims

1. A method for organizing spectrum map communication based on semantic information, characterized in that, The method is implemented in the mutual communication of a cloud / edge server and a user, and comprises: Obtaining a spectrum map and text information; Using a variational autoencoder to encode the spectrum map to obtain image feature information, and mapping the image feature information to a latent feature space to obtain an image latent space vector; Encoding the text information to obtain text features, and stacking the text features according to the attributes of the text prompt words in the same batch of feature dimensions to obtain a text prompt word feature vector; this step comprises: classifying the text prompt words in the text information according to a preset rule to obtain positive prompt words and negative prompt words, and stacking the positive prompt words and the negative prompt words in the same batch of feature dimensions to obtain a text prompt word feature vector; Based on a denoising diffusion implicit model, the image latent space vector is added with noise to obtain a noisy image feature; this step comprises: selecting a value same as the reconstruction ratio of the spectrum map as a noise value, adding noise to the image latent space vector based on the denoising diffusion implicit model to reconstruct the noise value percentage, and obtaining a noisy image feature corresponding to the noise percentage feature; After the noisy image feature and the text prompt word feature vector are fused, they are input into a UNet network together with a time step number for noise prediction, and the image latent space vector is optimized according to the predicted noise data to obtain a real spectrum map.

2. The method of claim 1, wherein, Encoding the text information to obtain text features comprises: Using a CLIP text vectorization of a CLIP text encoder to encode the text information to obtain encoded text information, inputting the encoded text information into a CLIP text model for feature extraction to obtain text features.

3. The method of claim 1, wherein, Setting a noise value and a total number of noise iteration steps, and adding noise to the image latent space vector according to the product of the noise value and the total number of noise iteration steps; Calculating the difference between the total number of noise iteration steps and the product to obtain a remaining iteration step number, and not adding noise to the image latent space vector in the remaining iteration step number.

4. The method of claim 3, wherein, After the noisy image feature and the text prompt word feature vector are fused, they are input into a UNet network together with a time step number for noise prediction, and the image latent space vector is optimized according to the predicted noise data to obtain a real spectrum map, comprising: The noisy image feature and the text prompt word feature vector are fused by a neural network model, and a time step number obtained by a fully connected layer of the neural network is input into a UNet network for noise prediction.

5. The method of claim 4, wherein, According to the predicted noise data, the image latent space vector is optimized to obtain a real spectrum map, comprising: After the UNet network predicts noise at each step, a new Gaussian noise is obtained, the Gaussian noise is input into the UNet network for denoising iteration according to a preset ratio, the image latent space vector is optimized by sampling in the latent space for multiple times, and a real spectrum map is reconstructed according to the optimized image latent space vector.

6. A device for organizing spectrum map communication based on semantic information, characterized by, The device is loaded in a mutual communication system of a cloud / edge server and a user, and is used to implement the method of any one of claims 1 to 5, and the device comprises: An information acquisition module is configured to acquire a spectrum map and text information. An image latent space vector acquisition module is configured to encode the spectrum map by using a variational autoencoder to obtain image feature information, map the image feature information to a latent feature space, and obtain an image latent space vector. A text prompt word feature vector acquisition module is configured to encode the text information to obtain text features, stack the text features according to the attributes of the text prompt words in the same batch of feature dimensions, and obtain a text prompt word feature vector. A noisy image feature acquisition module is configured to add noise to the image latent space vector based on a denoising diffusion implicit model to obtain a noisy image feature. A real spectrum map construction module is configured to fuse the noisy image feature and the text prompt word feature vector, input them to a UNet network together with a time step number for noise prediction, optimize the image latent space vector according to the predicted noise data, and obtain a real spectrum map. 7.A computer device, comprising a memory and a processor, wherein the memory stores a computer program, and the computer device is configured to perform the method according to any one of claims 1-6 when the computer program is executed by the processor. The processor executes the computer program to implement the steps of the method in any one of claims 1 to 5.

Images