A method and system for optimizing energy efficiency of decellularized network based on intelligent metasurface

Abstract

The present invention discloses a method and system for optimizing energy efficiency of a decellularized network based on an intelligent super surface, and relates to the technical field of network energy efficiency optimization, including deploying sensors to collect decellularized network and channel data and storing them in a database after preprocessing; constructing a decellularized network model based on database data; defining a decellularized network intelligent super surface configuration, and optimizing the intelligent super surface configuration based on a particle swarm optimization algorithm; and optimizing the energy efficiency of the decellularized network based on the intelligent super surface configuration. The present invention constructs a decellularized network model by collecting decellularized network related data, and optimizes the intelligent super surface configuration based on the decellularized network model and the particle swarm optimization algorithm, thereby completing the energy efficiency optimization of the decellularized network, effectively improving the operating energy efficiency and stability of the decellularized network, and greatly improving the applicability and flexibility of the decellularized network.

Description

Technical Field

[0001] The present invention relates to the technical field of network energy efficiency optimization, and in particular to a method and system for optimizing energy efficiency of a de-cellularized network based on an intelligent metasurface. Background Art

[0002] In the past few years, with the rapid development of mobile communication technology, decellularized networks, as an emerging communication architecture, have gradually attracted widespread attention from researchers and the industry. Decellularized networks are designed through distributed network architecture to improve network coverage, capacity and energy efficiency, especially in densely populated areas and complex geographical environments. With the advancement of 5G and future 6G technologies, the optimization of decellularized networks has become the key to improving network performance and user experience. Smart metasurface technology, as a wireless communication technology that has rapidly emerged in recent years, dynamically adjusts the reflection path of electromagnetic waves to achieve active control of the signal propagation environment, thereby optimizing the signal coverage and energy efficiency of the network, becoming one of the important technologies for decellularized network optimization. Existing technologies face many challenges in implementing energy efficiency optimization of decellularized networks. Existing network models and optimization algorithms often fail to fully consider the specific characteristics of decellularized networks, resulting in deviations between the optimization results and the actual application scenarios, and the optimization effect is also poor. Summary of the invention

[0003] In view of the problems existing in the above-mentioned existing de-cellular network energy efficiency optimization method and system based on intelligent metasurface, the present invention is proposed.

[0004] Therefore, the problem to be solved by the present invention is that the existing network models and optimization algorithms often fail to fully consider the specific characteristics of decellularized networks, resulting in a deviation between the optimization results and the actual application scenarios, and the optimization effect is also poor.

[0005] To solve the above technical problems, the present invention provides the following technical solutions: a method for optimizing the energy efficiency of a de-cellularized network based on an intelligent metasurface, which comprises deploying sensors to collect de-cellularized network and channel data and storing them in a database after pre-processing; constructing a de-cellularized network model based on database data; defining the de-cellularized network intelligent metasurface configuration, and optimizing the intelligent metasurface configuration based on a particle swarm optimization algorithm; and optimizing the energy efficiency of the de-cellularized network based on the intelligent metasurface configuration.

[0006] As a preferred solution of the de-cellular network energy efficiency optimization method based on intelligent metasurface described in the present invention, wherein: the deployment of sensors to collect de-cellular network and channel data and pre-processing refers to deploying loT sensors on de-cellular network devices to collect de-cellular network data and channel data and regularly sending them to the data processing center through a wireless network, and after the loT sensors are deployed, the sensor coverage is verified, the collected data is cleaned and filtered, duplicate values ​​are deleted, missing values ​​are filled and converted into a unified format, and feature engineering is used to extract feature vectors of the pre-processed data.

[0007] As a preferred solution of the de-cellular network energy efficiency optimization method based on intelligent metasurface described in the present invention, wherein: the storage in the database to the data processing center stores the data in the database after pre-processing the data, and adds security access rights and access passwords to the stored data. The database backs up the stored data and stores the backup data in the cloud, and regularly performs integrity checks on the stored data to generate a test report for synchronous storage.

[0008] As a preferred solution of the de-cellularized network energy efficiency optimization method based on intelligent metasurface described in the present invention, wherein: the de-cellularized network model constructed based on database data includes the following steps:

[0009] Use the preprocessed data to build a graph structure, take the entities in the decellularized network as nodes, and the connections between entities as edges. The decellularized network model formula is expressed as:

[0010]

[0011]

[0012] in To output the cellular network model, is the total number of nodes, and is the mean and standard deviation of the eigenvector, is the information filtering function, are model parameters, represents the set of neighbor nodes of node i, and Represent the feature vectors of nodes i and j respectively, and Represents the number of edges connected to nodes i and j, is the bias term, is a nonlinear activation function;

[0013] After building the decellularized network model, the model parameters are initialized and values ​​are randomly selected from a normal distribution with a mean of 0 and a standard deviation of 0.01 as model parameters. and the bias term The initial value of

[0014] Define the loss function to quantify the difference between the model output value and the actual value:

[0015]

[0016] in The loss quantified for the difference, is the actual label of node i, is the model output of node i;

[0017] Perform iterative training, calculate the model prediction output, use the loss function to calculate the gap between the model prediction and the actual label, and use the gradient descent algorithm to adjust the model parameters. and the bias term Iterative updates:

[0018]

[0019] in Represents the model parameters, including and , is the learning rate, is the loss function about The gradient of

[0020] The iteration stops when the model loss no longer decreases significantly or reaches the predetermined number of iterations, and the updated model parameters are output. and the bias term ;

[0021] Based on the updated model parameters and the bias term Calculate the attention coefficient:

[0022]

[0023] in is the attention coefficient of the i-th node to its neighbor j node, and are the weight vectors of the i-th node and the j-th node respectively;

[0024] Update node features by weighted aggregation of neighbor node feature vectors:

[0025]

[0026] in is the updated feature vector of node i, is the sigmoid function;

[0027] The updated node feature vector is input into the de-cellularized network model to complete the model optimization and obtain the final de-cellularized network model.

[0028] As a preferred solution of the decellularized network energy efficiency optimization method based on smart metasurface described in the present invention, wherein: the definition of the decellularized network smart metasurface configuration refers to defining the smart metasurface configuration based on the decellularized network model, and forming a set of smart metasurface configuration parameters:

[0029]

[0030] Each of the parameters represents the configuration parameters of the smart metasurface;

[0031] The intelligent hypersurface configuration reward function is defined as:

[0032]

[0033] in is the reward function output, representing the performance index of the smart metasurface, For smart metasurfaces in configuration The frequency response of the reflected wave, For smart metasurfaces in configuration The energy consumption under and is the weight parameter, and the smart hypersurface configuration is input into the reward function for performance evaluation.

[0034] As a preferred solution of the decellularized network energy efficiency optimization method based on the intelligent super surface of the present invention, the intelligent super surface configuration optimization based on the particle swarm optimization algorithm refers to the intelligent super surface configuration optimization using the particle swarm optimization algorithm after defining the intelligent super surface configuration:

[0035] Define a particle swarm, where each particle represents a smart metasurface configuration and the particle position vector is expressed as ,in represents the phase adjustment value of the smart metasurface configuration, and n is the number of smart metasurface configuration parameters;

[0036] Initialize the position vector of a group of particles in the particle swarm and the velocity vector is a random value, and the initialization range is:

[0037]

[0038]

[0039] in To generate random numbers, and are the maximum and minimum speeds of the particle motion respectively;

[0040] Iteratively update particle positions and velocities:

[0041]

[0042]

[0043] in is the inertia weight, and is the acceleration factor, is the individual historical optimal position vector of particle i, is the global optimal position of the particle swarm, t is the number of iterations;

[0044] After each iteration, the intelligent hypersurface configuration reward function is applied to the particle position vector to obtain the function output Until the iteration gets the maximum output Or the iteration is stopped after reaching a preset number of iterations, and the intelligent hypersurface configuration in the particle position vector is optimized and output.

[0045] As a preferred scheme of the de-cellularized network energy efficiency optimization method based on intelligent metasurface described in the present invention, wherein: after the de-cellularized network energy efficiency is optimized based on the intelligent metasurface configuration to obtain the optimized intelligent metasurface configuration, the intelligent metasurface RIS unit parameters are adjusted according to the configuration, the control channel is adjusted to complete the de-cellularized network energy efficiency optimization, and after the adjustment is completed, the de-cellularized network optimization effect is evaluated through the network testing tool.

[0046] Another object of the present invention is to provide a decellularized network energy efficiency optimization system based on intelligent metasurface, which comprises:

[0047] A data collection module is used to deploy sensors to collect cellular network data and channel data and perform pre-processing;

[0048] Network analysis module, used to build decellularized network models;

[0049] Configuration optimization module, used to optimize the configuration of the smart metasurface of the decellularized network and complete the energy efficiency optimization of the decellularized network;

[0050] The data storage module is used to store and securely protect the data collected and analyzed.

[0051] A computer device comprises: a memory and a processor; the memory stores a computer program, and the processor implements the steps of the above-mentioned method for optimizing energy efficiency of a de-cellular network based on a smart metasurface when executing the computer program.

[0052] A computer-readable storage medium stores a computer program, which, when executed by a processor, implements the steps of the above-mentioned method for optimizing energy efficiency of a de-cellular network based on a smart metasurface.

[0053] The beneficial effects of the present invention are as follows: the present invention constructs a de-cellularized network model by collecting de-cellularized network related data, and defines the intelligent metasurface configuration for optimization based on the de-cellularized network model and the particle swarm optimization algorithm, thereby completing the energy efficiency optimization of the de-cellularized network, effectively improving the operating energy efficiency and stability of the de-cellularized network, and greatly improving the applicability and flexibility of the de-cellularized network. BRIEF DESCRIPTION OF THE DRAWINGS

[0054] In order to more clearly illustrate the technical solutions of the embodiments of the present invention, the accompanying drawings required for use in the description of the embodiments will be briefly introduced below. Obviously, the accompanying drawings described below are only some embodiments of the present invention. For ordinary technicians in this field, other accompanying drawings can be obtained based on these accompanying drawings without paying creative work.

[0055] Figure 1 Schematic diagram of the flow of the decellularized network energy efficiency optimization method based on smart metasurface.

[0056] Figure 2 Schematic diagram of the structure of the de-cellular network energy efficiency optimization system based on smart metasurface. DETAILED DESCRIPTION

[0057] In order to make the above-mentioned objects, features and advantages of the present invention more obvious and easy to understand, the specific embodiments of the present invention are described in detail below in conjunction with the accompanying drawings.

[0058] In the following description, many specific details are set forth to facilitate a full understanding of the present invention, but the present invention may also be implemented in other ways different from those described herein, and those skilled in the art may make similar generalizations without violating the connotation of the present invention. Therefore, the present invention is not limited to the specific embodiments disclosed below.

[0059] 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 term "in one embodiment" that appears in different places in this specification does not necessarily refer to the same embodiment, nor is it a separate or selective embodiment that is mutually exclusive with other embodiments.

[0060] Example 1

[0061] Reference Figure 1, which is the first embodiment of the present invention, and this embodiment provides a method for optimizing energy efficiency of a decellularized network based on an intelligent super surface, and the method for optimizing energy efficiency of a decellularized network based on an intelligent super surface includes the following contents:

[0062] S1, deploy sensors to collect cellular network and channel data and store them in the database after preprocessing;

[0063] Specifically, deploying sensors to collect de-cellular network and channel data and pre-processing it means deploying loT sensors on de-cellular network devices to collect de-cellular network data and channel data and regularly sending them to a data processing center through a wireless network. After the loT sensors are deployed, the sensor coverage is verified, the collected data is cleaned and filtered, duplicate values ​​are deleted, missing values ​​are filled and converted into a unified format, and feature engineering is used to extract feature vectors of the pre-processed data.

[0064] By deploying loT sensors on decellular network devices, real-time monitoring of network environment and device status can be achieved, including the collection of multi-dimensional information such as user distribution, device performance and environmental factors, which provides a rich data foundation for dynamic optimization of the network. The collected data can be used for real-time monitoring of network status, evaluation of network performance, and timely discovery and resolution of problems, especially in areas with insufficient signal coverage or frequent changes. By cleaning and filtering the collected data, including deleting duplicate values, filling missing values ​​and converting them into a unified format, the quality and availability of the data can be significantly improved, and errors in subsequent analysis and processing can be reduced. The purified data is more suitable for in-depth analysis and model training, which improves the accuracy and efficiency of data-driven decision-making. Especially when performing network optimization and configuration adjustments, feature engineering can extract more representative and informative features from the raw data, enhancing the expressiveness of data analysis and machine learning models. Through carefully selected feature vectors, subtle differences in network status and environmental changes can be more accurately captured.

[0065] Furthermore, the data is stored in the database and the data processing center stores the data in the database after preprocessing the data, and adds security access rights and access passwords to the stored data. The database backs up the stored data and stores the backup data in the cloud. At the same time, the stored data is regularly checked for integrity, and a test report is generated and stored synchronously.

[0066] The preprocessed data is stored in the database, which can improve the availability and query efficiency of the data. At the same time, through structured organization, it provides convenience for subsequent data analysis and processing. The structured stored data can be used in various business scenarios, such as network performance analysis, user behavior research, fault diagnosis, etc., to provide data support for decision-making. By setting access permissions and passwords, the data in the database can be effectively protected from unauthorized access or malicious attacks, ensuring data security and user privacy. This measure is very critical to maintaining user trust and complying with data protection regulations. Especially when dealing with personal sensitive information and commercial secrets, backing up and storing data in the cloud can improve data security and disaster recovery capabilities. Regular data integrity testing can promptly detect data corruption or tampering problems and ensure data accuracy and integrity. Data backup and integrity testing are of great significance in ensuring business continuity, reducing the risk of data loss, and meeting compliance requirements. In the event of a network attack or system failure, data services can be quickly restored.

[0067] S2, building a decellularized network model based on database data;

[0068] Specifically, building a decellularized network model based on database data includes the following steps:

[0069] Use the preprocessed data to build a graph structure, take the entities in the decellularized network as nodes, and the connections between entities as edges. The decellularized network model formula is expressed as:

[0070]

[0071]

[0072] in To output the cellular network model, is the total number of nodes, and is the mean and standard deviation of the eigenvector, is the information filtering function, are model parameters, represents the set of neighbor nodes of node i, and Represent the feature vectors of nodes i and j respectively, and Represents the number of edges connected to nodes i and j, is the bias term, is a nonlinear activation function;

[0073] After building the decellularized network model, the model parameters are initialized and values ​​are randomly selected from a normal distribution with a mean of 0 and a standard deviation of 0.01 as model parameters. and the bias term The initial value of

[0074] Define the loss function to quantify the difference between the model output value and the actual value:

[0075]

[0076] in The loss quantified for the difference, is the actual label of node i, is the model output of node i;

[0077] Perform iterative training, calculate the model prediction output, use the loss function to calculate the gap between the model prediction and the actual label, and use the gradient descent algorithm to adjust the model parameters. and the bias term Iterative updates:

[0078]

[0079] in Represents the model parameters, including and , is the learning rate, is the loss function about The gradient of

[0080] The iteration stops when the model loss no longer decreases significantly or reaches the predetermined number of iterations, and the updated model parameters are output. and the bias term ;

[0081] Based on the updated model parameters and the bias term Calculate the attention coefficient:

[0082]

[0083] in is the attention coefficient of the i-th node to its neighbor j node, and are the weight vectors of the i-th node and the j-th node respectively;

[0084] Update node features by weighted aggregation of neighbor node feature vectors:

[0085]

[0086] in is the updated feature vector of node i, is the sigmoid function;

[0087] The updated node feature vector is input into the de-cellularized network model to complete the model optimization and obtain the final de-cellularized network model.

[0088] By representing the decellularized network as a graph structure, the complex relationships between entities in the network can be simulated more naturally, improving the accuracy and depth of network analysis and optimization. This method can be used in the network design stage to predict network performance and identify potential structural problems, providing a scientific basis for network planning and expansion. By randomly initializing model parameters from a normal distribution, a variety of starting points are provided for model training, which helps to avoid local optimality. Iterative training enables the model to gradually adapt to the data and optimize performance. This step is crucial to the accuracy and generalization ability of the model, ensuring that the model can be effectively applied to different network environments and conditions. By calculating the attention coefficient, the model can pay more attention to neighbor nodes that are more important to the prediction task, thereby improving the prediction accuracy and efficiency of the model. The attention mechanism makes the model more efficient in processing large-scale network data, especially in decellularized networks with large network scale and complex connections, which can significantly improve the quality and speed of information processing. Through repeated iterative optimization, the model is not only refined in theory, but also can show better performance and stability in practical applications. The optimized decellularized network model can provide strong support for network management and maintenance, including but not limited to traffic management, fault prediction, network security and other aspects.

[0089] S3, define the configuration of the decellularized intelligent super surface, and optimize the intelligent super surface configuration based on the particle swarm optimization algorithm;

[0090] Specifically, defining the decellularized intelligent supersurface configuration means defining the intelligent supersurface configuration based on the decellularized network model, and forming a set of intelligent supersurface configuration parameters:

[0091]

[0092] Each of the parameters represents the configuration parameters of the smart metasurface;

[0093] The intelligent hypersurface configuration reward function is defined as:

[0094]

[0095] in is the reward function output, representing the performance index of the smart metasurface, For smart metasurfaces in configuration The frequency response of the reflected wave, For smart metasurfaces in configuration The energy consumption under and is the weight parameter, and the smart hypersurface configuration is input into the reward function for performance evaluation.

[0096] By accurately modeling the decellularized network environment and defining a set of configuration parameters for the smart metasurface, fine control of the performance of the smart metasurface can be achieved. This method allows the configuration to be dynamically adjusted according to the actual needs of the network environment to optimize network performance. This step is crucial to improving the adaptability and performance of decellularized networks in complex environments, and is especially suitable for modern wireless communication scenarios with changing requirements, such as smart cities and autonomous driving vehicle communication networks. By introducing a reward function to evaluate the configuration performance of the smart metasurface, the effects of different configuration schemes can be quantified, providing a basis for selecting the optimal configuration. The reward function takes into account multiple factors such as performance index, frequency response, and energy consumption, making the performance The evaluation is more comprehensive and accurate. The definition of the reward function is extremely critical for the automatic configuration and optimization of the intelligent metasurface. It can guide the algorithm to find the configuration scheme with the best energy efficiency and performance, which is especially important for energy-constrained wireless network environments. It helps to extend the life of the equipment and improve the user experience. By inputting the configuration parameters of the intelligent metasurface into the reward function for performance evaluation, we can intuitively understand the specific impact of different configurations on network performance, which is convenient for quickly identifying the best configuration scheme. Performance evaluation can not only be used for the initial configuration optimization of the intelligent metasurface, but also for dynamic adjustments during network operation, which enables the network to respond to environmental changes in real time and maintain the best service quality.

[0097] Furthermore, the intelligent supersurface configuration optimization based on the particle swarm optimization algorithm refers to the intelligent supersurface configuration optimization using the particle swarm optimization algorithm after the intelligent supersurface configuration is defined:

[0098] Define a particle swarm, where each particle represents a smart metasurface configuration and the particle position vector is expressed as ,in represents the phase adjustment value of the smart metasurface configuration, and n is the number of smart metasurface configuration parameters;

[0099] Initialize the position vector of a group of particles in the particle swarm and the velocity vector is a random value, and the initialization range is:

[0100]

[0101]

[0102] in To generate random numbers, and are the maximum and minimum speeds of the particle motion respectively;

[0103] Iteratively update particle positions and velocities:

[0104]

[0105]

[0106] in is the inertia weight, and is the acceleration factor, is the individual historical optimal position vector of particle i, is the global optimal position of the particle swarm, t is the number of iterations;

[0107] After each iteration, the intelligent hypersurface configuration reward function is applied to the particle position vector to obtain the function output Until the iteration gets the maximum output Or the iteration is stopped after reaching a preset number of iterations, and the intelligent hypersurface configuration in the particle position vector is optimized and output.

[0108] The PSO algorithm simulates the social behavior of particle groups to find the optimal solution to the problem, so that the intelligent metasurface configuration optimization process does not depend on the gradient information of the problem. It is suitable for nonlinear and non-convex optimization problems. The method is simple, efficient, easy to implement, and can quickly converge to the global optimal or approximate optimal solution. The application of the PSO algorithm in the optimization of intelligent metasurface configuration can provide better signal coverage and network performance for wireless communication networks. It is particularly suitable for signal enhancement and interference management in complex environments. By optimizing the configuration parameters of the intelligent metasurface, such as the phase adjustment value, the reflection characteristics of the wireless signal can be significantly improved, and fine control of the signal propagation environment can be achieved. This control can enhance signal strength, expand coverage, and reduce interference. , thereby improving the overall performance and energy efficiency of the network. The optimized smart metasurface configuration can be widely used in 5G and future 6G networks, the Internet of Things (loT), intelligent transportation systems and other fields, especially in dense urban environments and remote areas that require high-quality wireless connections. By iteratively updating the position and velocity of particles, the algorithm can effectively search in the solution space while avoiding falling into local optimal solutions. The design of the acceleration coefficient and inertia weight enables the algorithm to have the ability to self-adjust and dynamically adjust the search strategy based on the experience in the search process. This iterative update strategy makes the PSO algorithm very suitable for dealing with dynamically changing optimization problems, such as real-time adjustment of the configuration of smart metasurfaces in mobile communication networks to cope with changes in network load and user movement.

[0109] S4. Optimize the energy efficiency of decellularized networks based on intelligent metasurface configuration.

[0110] Specifically, the de-cellularized network energy efficiency is optimized based on the intelligent metasurface configuration. After the optimized intelligent metasurface configuration is obtained, the intelligent metasurface RIS unit parameters are adjusted according to the configuration, the control channel is adjusted to complete the de-cellularized network energy efficiency optimization, and after the adjustment is completed, the de-cellularized network optimization effect is evaluated through the network testing tool.

[0111] By precisely adjusting the unit parameters of the smart metasurface, it is possible to dynamically control the wireless signal path and optimize the signal propagation conditions, thereby improving signal coverage and quality without increasing additional transmission power. This adjustment is particularly suitable for areas with poor indoor and outdoor signals, high-density network environments, and occasions with strict requirements on energy efficiency, such as smart buildings, large commercial centers, and smart cities. After adjusting the configuration of the smart metasurface, the energy efficiency of the decellularized network has been significantly improved. The optimized network not only maintains the service quality while consuming less energy, but also reduces signal attenuation and interference due to the optimization of the signal path. The improvement of energy efficiency is of great significance for reducing the operator's operation and maintenance costs, extending the life of equipment, and supporting sustainable development goals. By evaluating the optimized decellularized network through network testing tools, the impact of the optimization of the smart metasurface configuration on network performance can be quantified, including improvements in indicators such as signal coverage, communication quality, and network capacity.

[0112] Example 2

[0113] Reference Figure 2 , which is the second embodiment of the present invention, and which is different from the previous embodiment, provides a decellularized network energy efficiency optimization system based on an intelligent metasurface, which includes:

[0114] A data collection module is used to deploy sensors to collect cellular network data and channel data and perform pre-processing;

[0115] Network analysis module, used to build decellularized network models;

[0116] Configuration optimization module, used to optimize the configuration of the smart metasurface of the decellularized network and complete the energy efficiency optimization of the decellularized network;

[0117] The data storage module is used to store and securely protect the data collected and analyzed.

[0118] If the functions are implemented in the form of software functional units and sold or used as independent products, they can be stored in a computer-readable storage medium. Based on this understanding, the technical solution of the present invention, or the part that contributes to the prior art, or the part of the technical solution, can be embodied in the form of a software product. The computer software product is stored in a storage medium, including several instructions to enable a computer device (which can be a personal computer, server, or network device, etc.) to perform all or part of the steps of the methods described in various embodiments of the present invention. The aforementioned storage medium includes: U disk, mobile hard disk, read-only memory (ROM, Read-Only Memory), random access memory (RAM, Random Access Memory), disk or optical disk, etc., which can store program codes.

[0119] The logic and / or steps represented in the flowchart or otherwise described herein, for example, can be considered as an ordered list of executable instructions for implementing logical functions, and can be embodied in any computer-readable medium for use by an instruction execution system, device or apparatus (such as a computer-based system, a system including a processor, or other system that can fetch instructions from an instruction execution system, device or apparatus and execute instructions), or in conjunction with such instruction execution systems, devices or apparatuses. For the purposes of this specification, "computer-readable medium" can be any device that can contain, store, communicate, propagate or transmit a program for use by an instruction execution system, device or apparatus, or in conjunction with such instruction execution systems, devices or apparatuses.

[0120] More specific examples of computer-readable media (a non-exhaustive list) include the following: an electrical connection with one or more wires (electronic device), a portable computer disk case (magnetic device), a random access memory (RAM), a read-only memory (ROM), an erasable and programmable read-only memory (EPROM or flash memory), an optical fiber device, and a portable compact disk read-only memory (CDROM). In addition, the computer-readable medium may even be a paper or other suitable medium on which the program is printed, since the program may be obtained electronically, for example, by optically scanning the paper or other medium, followed by editing, deciphering or, if necessary, processing in another suitable manner, and then stored in a computer memory.

[0121] It should be understood that the various parts of the present invention can be implemented by hardware, software, firmware or a combination thereof. In the above embodiments, multiple steps or methods can be implemented by software or firmware stored in a memory and executed by a suitable instruction execution system. For example, if implemented by hardware, as in another embodiment, it can be implemented by any one of the following technologies known in the art or their combination: a discrete logic circuit with a logic gate circuit for implementing a logic function for a data signal, a dedicated integrated circuit with a suitable combination of logic gate circuits, a programmable gate array (PGA), a field programmable gate array (FPGA), etc.

[0122] It should be noted that the above embodiments are only used to illustrate the technical solutions of the present invention rather than to limit it. Although the present invention has been described in detail with reference to the preferred embodiments, those skilled in the art should understand that the technical solutions of the present invention may be modified or replaced by equivalents without departing from the spirit and scope of the technical solutions of the present invention, which should all be included in the scope of the claims of the present invention.

Claims

1. A method for optimizing energy efficiency of a decellularized network based on an intelligent metasurface, characterized in that: include, Deploy sensors to collect cellular network and channel data and store them in the database after preprocessing; Build a decellularized network model based on database data; Define the configuration of the decellularized intelligent metasurface and optimize it based on the particle swarm optimization algorithm; Decellularized network energy efficiency optimization based on intelligent metasurface configuration; Deploying sensors to collect and pre-process decellularized network and channel data refers to deploying loT sensors on decellularized network devices to collect decellularized network data and channel data and sending them to a data processing center via a wireless network on a regular basis, verifying the coverage of the sensors after deploying the loT sensors, cleaning and filtering the collected data, deleting duplicate values, filling in missing values ​​and converting them into a unified format, and extracting feature vectors of the pre-processed data using feature engineering; The storage in the database means that the data processing center stores the data in the database after preprocessing the data, and adds security access rights and access passwords to the stored data. The database backs up the stored data and stores the backup data in the cloud, and regularly performs integrity checks on the stored data to generate test reports for synchronous storage; The method of constructing a decellularized network model based on database data comprises the following steps: Use the preprocessed data to build a graph structure, take the entities in the decellularized network as nodes, and the connections between entities as edges. The decellularized network model formula is expressed as: Where G is the output of the decellularized network model, N is the total number of nodes, μ and σ are the mean and standard deviation of the feature vector, and f(x i , W) is the information filtering function, W is the model parameter, N(i) represents the neighbor node set of node i, x i and x j Represent the feature vectors of nodes i and j respectively, d i and d j Respectively represent the number of edges connected to nodes i and j, b is the bias term, and ReLU is the nonlinear activation function; After building the decellularized network model, the model parameters are initialized by randomly extracting values ​​from a normal distribution with a mean of 0 and a standard deviation of 0.01 as the initial values ​​of the model parameter W and the bias term b; Define the loss function to quantify the difference between the model output value and the actual value: Where L is the difference quantization loss, G i is the actual label of node i, is the model output of node i; Perform iterative training, calculate the model prediction output, use the loss function to calculate the gap between the model prediction and the actual label, and use the gradient descent algorithm to iteratively update the model parameters W and the bias term b: Where θ represents the model parameters, including W and b, α is the learning rate, is the gradient of the loss function with respect to θ; The iteration stops when the model loss no longer decreases significantly or reaches the predetermined number of iterations, and the updated model parameters W and bias term b are output; The attention coefficient is calculated based on the updated model parameters W and the bias term b: where a ij is the attention coefficient of the i-th node to its neighbor j node, c i and c j are the weight vectors of the i-th node and the j-th node respectively; Update node features by weighted aggregation of neighbor node feature vectors: where x i '' is the updated feature vector of node i, ρ is the sigmoid function; The updated node feature vector is input into the de-cellularized network model to complete the model optimization and obtain the final de-cellularized network model; Defining the decellularized network intelligent super surface configuration refers to defining the intelligent super surface configuration based on the decellularized network model, and forming a set of intelligent super surface configuration parameters: p={p1,p2,……,p n } Each parameter p n represents the configuration parameters of the smart metasurface; The intelligent hypersurface configuration reward function is defined as: R=ω*(∫|F {p} | 2 df)-β*E {p} Where R is the reward function output, represents the performance index of the smart metasurface, and F {p} is the frequency response of the reflected wave of the smart metasurface in configuration p, E {p} is the energy consumption of the smart hypersurface under configuration p, ω and β are weight parameters, and the smart hypersurface configuration is input into the reward function for performance evaluation; The de-cellular network energy efficiency optimization based on the intelligent metasurface configuration refers to adjusting the intelligent metasurface IRS unit parameters according to the configuration after obtaining the optimized intelligent metasurface configuration, controlling the channel adjustment to complete the de-cellular network energy efficiency optimization, and evaluating the de-cellular network optimization effect through the network testing tool after the adjustment is completed.

2. The method for optimizing energy efficiency of a decellularized network based on a smart metasurface according to claim 1, characterized in that: The intelligent super surface configuration optimization based on the particle swarm optimization algorithm refers to the intelligent super surface configuration optimization using the particle swarm optimization algorithm after the intelligent super surface configuration is defined: Define a particle swarm, where each particle represents a smart metasurface configuration and the particle position vector is represented by z i ={z i1 , z i2 , ..., z in }, where z in represents the phase adjustment value of the smart metasurface configuration, and n is the number of smart metasurface configuration parameters; Initialize the position vector z of a group of particles in the particle swarm i and the velocity vector v i is a random value, and the initialization range is: z i (0)=rand[0,2π] v i (0)=rand[V min ,V max ] Where rand is the generated random number, V max and V min are the maximum and minimum speeds of the particle motion respectively; Iteratively update particle positions and velocities: z i (t+1)=z i (t)+v i (t+1) in is the inertia weight, c1 and c2 are the acceleration coefficients, q best，i is the individual historical optimal position vector of particle i, g best is the global optimal position of the particle swarm, t is the number of iterations; After each iteration, the intelligent hypersurface configuration reward function is applied to the particle position vector to obtain the function output R until the iteration obtains the maximum output R or reaches the preset number of iterations, and the iteration is stopped to optimize the intelligent hypersurface configuration in the particle position vector and output it.

3. A de-cellular network energy efficiency optimization system based on a smart metasurface according to any one of claims 1 and 2, characterized in that: include, A data collection module is used to deploy sensors to collect cellular network data and channel data and perform pre-processing; Network analysis module, used to build decellularized network models; Configuration optimization module, used to optimize the configuration of the smart metasurface of the decellularized network and complete the energy efficiency optimization of the decellularized network; The data storage module is used to store and securely protect the data collected and analyzed.

4. A computer device comprising: Memory and processor; The memory stores a computer program, characterized in that: when the processor executes the computer program, the steps of the de-cellular network energy efficiency optimization method based on smart metasurface described in any one of claims 1 and 2 are implemented.

5. A computer-readable storage medium having a computer program stored thereon, characterized in that: When the computer program is executed by a processor, the steps of the de-cellularized network energy efficiency optimization method based on smart metasurface described in any one of claims 1 and 2 are implemented.

Images