Feature factor-based network dynamic planning method for edge Internet of Things equipment

Abstract

The invention discloses a characteristic factor-based network dynamic planning method for edge Internet of Things equipment. The method comprises the following steps of: S1, extracting characteristic factors and matching weights; s2, constructing a dynamic planning feature engineering model; s3, training a dynamic planning feature engineering model; and S4, optimizing the dynamic programming feature engineering model. According to the invention, by extracting and analyzing the feature factors composed of the service scene features, the distribution strategy of the network resources is adjusted in real time according to different scene requirements, and the service requirements in the scene are closely combined with network resource management, so that the edge equipment can flexibly adjust the network strategy according to the scene change, and the network resource utilization rate is improved. The utilization efficiency of network resources is improved, the security of data transmission is enhanced, and the communication delay is reduced.

Description

Technical Field

[0001] The present invention relates to the field of Internet of Things, and in particular to a network dynamic planning method based on characteristic factors for edge Internet of Things devices. Background Art

[0002] With the rapid development of Internet of Things (IoT) technology, more and more edge devices are widely used in various scenarios, such as smart cities, smart healthcare, industrial IoT, etc. Edge devices in these scenarios usually need to process a large amount of business data and transmit it efficiently through the network. However, the business characteristics in different scenarios vary, and the network requirements and optimization goals faced by edge devices are also different. Traditional network planning methods are usually configured based on network conditions and approximate existing resources. They lack fine-grained perception of business data and cannot dynamically adjust network resource configuration according to actual business needs. In this context, traditional static network planning methods cannot effectively cope with the dynamic and changing IoT environment.

[0003] The complexity of IoT scenarios is not only reflected in the increase in the number of devices and the amount of data, but also in the fact that different business scenarios have strong differences in their demands for network resources. For example, in smart medical scenarios, sensitive data involving personal privacy requires encrypted channels for transmission; while in smart factories, high-concurrency production control data requires low latency and high bandwidth guarantees. Such scenario-specific requirements make it difficult for a unified network planning strategy to meet the real-time needs of different business scenarios.

[0004] To solve this problem, existing network management methods attempt to introduce technologies such as machine learning to predict and optimize network status. However, these methods often rely on a large amount of historical data and pre-trained models, making it difficult to flexibly respond to the rapidly changing IoT environment. In addition, they also lack sufficient consideration of scenario business characteristics and cannot perform personalized allocation of network resources based on specific business needs. Summary of the invention

[0005] The present invention provides a network dynamic planning method for edge IoT devices based on characteristic factors, which closely combines business needs in different scenarios with network resource management to improve network resource utilization efficiency, enhance network security, and reduce communication delays. It can solve at least one of the above technical problems.

[0006] In order to solve the above technical problems, the present invention adopts the following technical solution: a network dynamic planning method of edge IoT devices based on characteristic factors, comprising the following steps:

[0007] S1. Extraction and weighting of characteristic factors: Extract general characteristic factors and service characteristic factors, and assign weights to the extracted characteristic factors to assist in scheduling network resources.

[0008] S2. Constructing a dynamic programming feature engineering model: constructing the dynamic programming feature engineering model based on the feature factors obtained in S1, wherein the dynamic programming feature engineering model uses a reinforcement learning model deep Q network to process network dynamic programming problems, adjust and output the optimal network resource allocation strategy, and dynamically adapt network resources to business needs;

[0009] S3, training a dynamic programming feature engineering model: based on the dynamic programming feature engineering model obtained in S2, at least one or more of the experience replay mechanism, Q value update or target network is used to perform model training;

[0010] S4. Optimize the dynamic programming feature engineering model: update the feature factor data input into the dynamic programming feature engineering model in real time. The dynamic programming feature engineering model re-plans network resource allocation when the network status changes or the business demand fluctuates, and gradually optimizes and updates during operation.

[0011] Further, in S1, the general characteristic factors are used to characterize general service indicators in general service scenarios, including at least a concurrent bandwidth adjustment factor, a response delay adjustment factor, and a connection stability factor;

[0012] The concurrent bandwidth adjustment factor is composed of the number of concurrent business services, the current bandwidth occupancy rate and the service request rate. When the system detects that the number of concurrent business services has increased, combined with the current bandwidth occupancy rate and the service request rate, the system evaluates and allocates more bandwidth resources to high-concurrency services in real time to ensure smooth data transmission and avoid congestion and delay.

[0013] The response delay adjustment factor is combined with the business service response time, the current network delay and the current load. When the business service response time exceeds the set threshold, the system extracts the characteristics of the current network delay and the current load, determines whether it is necessary to adjust the routing path or increase bandwidth resources, optimizes the network configuration, and reduces the delay of service response;

[0014] The connection stability factor is used to monitor the stability of the network and the reliability of the service connection. It is composed of the business service connection loss rate, data packet loss rate and network transmission error rate. When the business service connection loss rate or data packet loss rate is frequent, the system detects and adjusts the network's transmission strategy to ensure the stable operation of key businesses.

[0015] Furthermore, in S1, the business-specific diagnostic factors are used to characterize specific business indicators in specific business scenarios, including at least a data sensitivity factor, a real-time demand factor, and a concurrent number factor of standby communication;

[0016] The data sensitivity factor is used to determine whether the data needs to be transmitted through an encrypted channel;

[0017] The real-time requirement factor is used to control the low-delay requirement of the signal to ensure the instant response of the business operation;

[0018] The concurrency factor of the communication between the backup devices is used to determine the dynamic allocation demand of the bandwidth.

[0019] Furthermore, the specific business scenarios represented by the business-specific diagnosis factors include at least smart traffic scenarios, smart medical scenarios and smart factory scenarios, and the corresponding characteristic factors are traffic flow adjustment factors, medical data sensitivity factors and production load adjustment factors respectively;

[0020] The traffic flow adjustment factor is composed of real-time traffic flow, vehicle density and camera video data frame rate. In the intelligent traffic scenario, when the traffic flow increases, the system extracts the factor to dynamically adjust the bandwidth resources. The vehicle density and camera video data frame rate jointly determine the demand for bandwidth. The system ensures the real-time and clarity of the video stream by increasing bandwidth allocation, avoiding delays or freezes in the monitoring screen due to insufficient bandwidth.

[0021] The medical data sensitivity factor is composed of the patient data sensitivity level, real-time monitoring requirements and data transmission index requirements. In the smart medical scenario, the system extracts the factor to perform sensitivity assessment on the patient data. When the system detects that the patient data sensitivity level is highly sensitive data, the medical data sensitivity factor triggers the encryption transmission strategy and, based on the real-time monitoring requirements, prioritizes the allocation of high-bandwidth and low-latency network resources to ensure that the data transmission index requirements are met.

[0022] The production load adjustment factor is combined with the production line equipment load rate, real-time production data volume and operation task priority. In the smart factory scenario, when the production line equipment load rate increases, the system extracts the factor to evaluate the real-time load status and real-time production data volume of the production line equipment, ensuring that high-level critical tasks in the operation task priority meet the network resource requirements of high bandwidth and low latency.

[0023] Furthermore, in S1, the weight ratio of the characteristic factors is optimized by a weighted linear model of a multi-objective optimization algorithm based on historical data.

[0024] Furthermore, the S2 further comprises:

[0025] S21, data cleaning: before constructing the dynamic programming feature engineering model, preprocess the extracted feature factors to ensure the accuracy and consistency of the data;

[0026] Data cleaning operations at least include missing data processing, abnormal data filtering and duplicate data removal, among which missing data processing is to fill the missing data with the mean or median, and abnormal data filtering and duplicate data removal are to identify and filter outlier data through z-score or clustering-based methods;

[0027] S22, model construction: The core of the dynamic programming feature engineering model is network dynamic programming based on the feature factors, and the reinforcement learning model deep Q network is used to process the network dynamic programming problem. The network dynamic programming includes state space definition, action space definition and reward mechanism design;

[0028] The dynamic programming feature engineering model outputs an optimal network resource allocation strategy based on the input feature factors, and the network resource allocation strategy includes bandwidth allocation, routing path selection and transmission strategy adjustment.

[0029] Further, in the S22, in the scenario of network dynamic planning, the state space is composed of a plurality of the characteristic factors including the general characteristic factors and the service characteristic factors, and the specific state can be represented as a vector, which at least includes the current network status, network resource allocation status and service demand information;

[0030] The action space defines the various network resource allocation strategies that the system can adopt, that is, for each state, the reinforcement learning model deep Q network needs to select a strategy in the network resource allocation strategy and assign an action;

[0031] As the core part of network dynamic programming, the reward mechanism defines the feedback brought by each action. In the scenario of network dynamic programming, the reward mechanism is designed according to the following indicators:

[0032] ① If the selected action improves the service response speed, reduces latency, or reduces packet loss rate, a positive reward will be given;

[0033] ② If the selected action optimizes bandwidth utilization or reduces resource waste, a positive reward is given;

[0034] ③. In the transmission of sensitive data, additional rewards can be obtained by ensuring the secure transmission of data.

[0035] Further, in S22, the bandwidth allocation refers to dynamically adjusting the bandwidth allocation strategy between edge devices according to the bandwidth requirements of each service in the network, and giving priority to increasing bandwidth allocation for key tasks;

[0036] The routing path selection refers to selecting a low-latency or high-security routing path according to the weight of the characteristic factor to ensure that data is transmitted in an optimal manner;

[0037] The transmission strategy refers to deciding whether to enable encrypted transmission, or to select other appropriate transmission methods to ensure data security and transmission efficiency.

[0038] Furthermore, in S3, the experience replay mechanism means that after each action is executed, the system stores the tuple of state, action, reward and next state in the experience replay pool;

[0039] Q-value updating means randomly extracting a batch of data from the experience replay pool for training each time, and updating the Q-value function through the Q-learning algorithm;

[0040] The target network refers to the two neural networks used by the deep Q network of the reinforcement learning model. One is the main network, which is responsible for the current Q value prediction, and the other is the target network, which is used for regular updates to stabilize the training process.

[0041] Furthermore, the S4 further comprises:

[0042] S41, real-time monitoring: the system collects and obtains the characteristic factor data of the current network in real time through sensors or monitoring tools deployed on edge devices, and inputs the data into the dynamic programming feature engineering model;

[0043] S42, real-time decision-making: After the dynamic programming feature engineering model is fully trained, the reinforcement learning model deep Q network receives the current state information in real time and selects the optimal action based on the learned strategy. When the system detects changes in network state or fluctuations in business demand, the reinforcement learning model deep Q network quickly adjusts network resource allocation to adapt to the new environment;

[0044] S43, self-update: when a significant change in the network environment or business demand is detected, the system will automatically re-plan the network resource allocation according to the new feature factor weights, and gradually optimize the dynamic programming feature engineering model during operation through the online learning algorithm adaptive gradient method;

[0045] S44, Feedback mechanism: The system compares the results of each network planning with the expected goals, and continuously adjusts the model parameters through the feedback mechanism to ensure that the allocation of network resources is always optimal.

[0046] The beneficial effects of the present invention are embodied in:

[0047] 1. Improve network resource utilization: By dynamically adjusting bandwidth, routing and transmission strategies, the present invention can reasonably allocate network resources according to actual business needs, maximize network utilization efficiency, and reduce congestion and delay.

[0048] 2. Enhance network security: The present invention can automatically select an encrypted channel for data transmission by identifying sensitive information in business data, thereby ensuring the security of data during transmission. It is particularly suitable for scenarios that require privacy protection, such as medical and financial.

[0049] 3. Reduce communication delay and improve service quality: For business scenarios with high concurrency or high real-time requirements, the present invention can dynamically allocate bandwidth and optimize routing according to characteristic factors to ensure low-latency transmission of key data, thereby improving user experience and business service quality.

[0050] 4. Strong versatility and adaptability: This method is applicable to different IoT scenarios and edge devices, and has high versatility and adaptability. Through real-time perception of service characteristics and flexible network planning, the present invention can adapt to complex and changing IoT environments.

[0051] 5. Reduce energy consumption and computing overhead: By accurately extracting and selecting key characteristic factors, the present invention reduces the computational complexity and communication overhead of the system while ensuring network performance, and is suitable for resource-constrained edge devices. BRIEF DESCRIPTION OF THE DRAWINGS

[0052] The drawings described herein are used to provide further understanding of the present application and constitute a part of the present application. The illustrative embodiments of the present application and their descriptions are used to explain the present application and do not constitute improper limitations on the present application.

[0053] Figure 1 It is a flowchart of a network dynamic planning method according to an embodiment of the present invention.

[0054] Figure 2 It is a flowchart of the operation process of the network dynamic planning method according to an embodiment of the present invention.

[0055] Figure 3 It is a structural block diagram of a computer device according to an embodiment of the present invention. DETAILED DESCRIPTION

[0056] The technical solutions in the embodiments of the present invention will be clearly and completely described below in conjunction with the drawings in the embodiments of the present invention. Obviously, the described embodiments are only part of the embodiments of the present invention, rather than all of the embodiments. In the absence of conflict, the embodiments in this application and the features in the embodiments can be combined with each other. Based on the embodiments in the present invention, all other embodiments obtained by ordinary technicians in this field without making creative work are within the scope of protection of the present invention.

[0057] It should be noted that the meaning of “and / or” appearing in the full text includes three parallel solutions. Taking “A and / or B” as an example, it includes Solution A, Solution B, or a solution in which both A and B are satisfied.

[0058] It should be noted that those skilled in the art can understand that all or part of the steps implemented in the embodiments of the present invention can be implemented in whole or in part by software, hardware, firmware or any combination thereof. When implemented using hardware, it can be implemented in whole or in part in the form of purchasing standard parts or modified parts. When implemented using software, it can be implemented in whole or in part in the form of a computer program product. The computer program product includes one or more computer instructions. When the computer program instructions are loaded and executed on a computer, the process or function described in the embodiment of the present application is generated in whole or in part. The computer can be a general-purpose computer, a special-purpose computer, a computer network, or other programmable device. The computer instructions can be stored in a computer-readable storage medium, or transmitted from one computer-readable storage medium to another computer-readable storage medium. For example, the computer instructions can be transmitted from a website site, computer, server or data center by wired (e.g., coaxial cable, optical fiber, digital subscriber line (DSL)) or wireless (e.g., infrared, wireless, microwave, etc.) mode to another website site, computer, server or data center. The computer-readable storage medium can be any available medium that a computer can access or a data storage device such as a server or data center that includes one or more available media integrated. The available medium may be a magnetic medium (eg, a floppy disk, a hard disk, a magnetic tape), an optical medium (eg, a DVD), or a semiconductor medium (eg, a solid state disk (SSD)).

[0059] See also Figure 1-Figure 2 The embodiment of the present invention provides a method for dynamic network planning of edge IoT devices based on characteristic factors, comprising the following steps:

[0060] S1. Extraction and weighting of characteristic factors: Extract general characteristic factors and service characteristic factors, and assign weights to the extracted characteristic factors to assist in scheduling network resources.

[0061] S2. Constructing a dynamic programming feature engineering model: constructing the dynamic programming feature engineering model based on the feature factors obtained in S1, wherein the dynamic programming feature engineering model uses a reinforcement learning model deep Q network to process network dynamic programming problems, adjust and output the optimal network resource allocation strategy, and dynamically adapt network resources to business needs;

[0062] S3, training a dynamic programming feature engineering model: based on the dynamic programming feature engineering model obtained in S2, at least one or more of the experience replay mechanism, Q value update or target network is used to perform model training;

[0063] S4. Optimize the dynamic programming feature engineering model: update the feature factor data input into the dynamic programming feature engineering model in real time. The dynamic programming feature engineering model re-plans network resource allocation when the network status changes or the business demand fluctuates, and gradually optimizes and updates during operation.

[0064] The present invention provides a method for dynamic network planning based on characteristic factors. First, according to the business characteristics in different Internet of Things scenarios, core general characteristic factors and business characteristic factors are extracted. These factors can dynamically reflect the business needs of the current scenario through weight allocation ratios, and guide the system to perform network planning and resource allocation according to specific business needs. Secondly, through real-time analysis of the business scenario characteristic factors, the system allocates network resources according to a dynamic adjustment mechanism, such as bandwidth allocation, routing selection, and transmission channel encryption. In this way, the planning method of the present invention can run efficiently on edge devices with limited resources, and has strong adaptive capabilities. It can adjust the network planning in real time according to changes in network status and business needs, ensuring that the system can maintain efficient and stable operation in a dynamic Internet of Things environment.

[0065] In this embodiment, in S1, the general characteristic factors are used to characterize general service indicators in general service scenarios, including at least a concurrent bandwidth adjustment factor, a response delay adjustment factor, and a connection stability factor;

[0066] The concurrent bandwidth adjustment factor is composed of the number of concurrent business services, the current bandwidth occupancy rate and the service request rate. When the system detects that the number of concurrent business services has increased, combined with the current bandwidth occupancy rate and the service request rate, the system evaluates and allocates more bandwidth resources to high-concurrency services in real time to ensure smooth data transmission and avoid congestion and delay.

[0067] The response delay adjustment factor is combined with the business service response time, the current network delay and the current load. When the business service response time exceeds the set threshold, the system extracts the characteristics of the current network delay and the current load, determines whether it is necessary to adjust the routing path or increase bandwidth resources, optimizes the network configuration, and reduces the delay of service response;

[0068] The connection stability factor is used to monitor the stability of the network and the reliability of the service connection. It is composed of the business service connection loss rate, the data packet loss rate and the network transmission error rate. When the business service connection loss rate or the data packet loss rate is frequent, the system detects and adjusts the network's transmission strategy, such as rerouting, adding redundant paths or increasing connection priority, to ensure the stable operation of key businesses.

[0069] In this application, the universal characteristic factors can be applied to network resource requirements in multiple scenarios, such as network delay, bandwidth occupancy, device load, packet loss rate, etc. These characteristics are usually used as basic indicators to measure network performance and have high versatility in all scenarios. By monitoring these indicators in real time, the system can grasp the basic operating status of the network.

[0070] In this embodiment, in S1, the business-specific diagnostic factors are used to characterize specific business indicators in specific business scenarios, and at least include a data sensitivity factor, a real-time demand factor, and a concurrent number factor of standby communication;

[0071] The data sensitivity factor is used to determine whether the data needs to be transmitted through an encrypted channel;

[0072] The real-time requirement factor is used to control the low-delay requirement of the signal to ensure the instant response of the business operation;

[0073] The concurrency factor of the standby communication is used to determine the dynamic allocation demand of bandwidth;

[0074] The specific business scenarios represented by the business-specific diagnosis factors include at least smart traffic scenarios, smart medical scenarios and smart factory scenarios, and the corresponding characteristic factors are traffic flow adjustment factors, medical data sensitivity factors and production load adjustment factors respectively;

[0075] The traffic flow adjustment factor is composed of real-time traffic flow, vehicle density and camera video data frame rate. In the intelligent traffic scenario, when the traffic flow increases, the system extracts the factor to dynamically adjust the bandwidth resources. The vehicle density and camera video data frame rate jointly determine the demand for bandwidth. The system ensures the real-time and clarity of the video stream by increasing bandwidth allocation, avoiding delays or freezes in the monitoring screen due to insufficient bandwidth.

[0076] The medical data sensitivity factor is composed of the patient data sensitivity level, real-time monitoring requirements and data transmission index requirements. In the smart medical scenario, the system extracts the factor to perform sensitivity assessment on the patient data. When the system detects that the patient data sensitivity level is highly sensitive data (such as personal health data or surgical information), the medical data sensitivity factor triggers the encryption transmission strategy and, based on the real-time monitoring requirements (such as remote surgery or health monitoring), prioritizes the allocation of high-bandwidth and low-latency network resources to ensure that the data transmission index requirements are met;

[0077] The production load adjustment factor is combined with the production line equipment load rate, real-time production data volume and operation task priority. In the smart factory scenario, when the production line equipment load rate increases, the system extracts this factor to evaluate the real-time load status and real-time production data volume of the production line equipment, ensuring that the high-level critical tasks in the operation task priority meet the network resource requirements of high bandwidth and low latency. The system can automatically adjust network resources to ensure high-speed communication between production line equipment to avoid production interruptions or delays.

[0078] In this application, the business characteristic factors are different from the above-mentioned general characteristic factors. They are extracted according to the needs of specific business scenarios. Common specific business scenarios include the smart transportation scenarios, smart medical scenarios and smart factory scenarios mentioned above, and should also include other business scenarios, which are not listed one by one in this embodiment. These business characteristic factors change dynamically according to the needs of different business scenarios and determine the specific planning strategy of the network.

[0079] In this application, the extraction of the characteristic factors is the basis of network dynamic planning, ensuring that the system can make decisions based on different characteristic factors in business scenarios.

[0080] See also Figure 2 In this embodiment, in S1, the weight ratio of the characteristic factors is optimized by a weighted linear model of a multi-objective optimization algorithm based on historical data. The extracted general characteristic factors and the business characteristic factors need to be weighted to reflect their relative importance to network resource scheduling. The weight ratio is optimized differently according to different business scenarios. For example, in scenarios with high real-time requirements, the weight of the delay feature will be higher than the bandwidth utilization, while in scenarios with high data sensitivity, the weight of encrypted transmission will be increased.

[0081] See also Figure 1-Figure 2 In this embodiment, S2 further includes:

[0082] S21, data cleaning: before constructing the dynamic programming feature engineering model, preprocess the extracted feature factors to ensure the accuracy and consistency of the data;

[0083] Data cleaning operations at least include missing data processing, abnormal data filtering and duplicate data removal, among which missing data processing is to fill the missing data with the mean or median, and abnormal data filtering and duplicate data removal are to identify and filter outlier data through z-score or clustering-based methods;

[0084] S22. Model construction: The core of the dynamic programming feature engineering model is the network dynamic programming based on the feature factors. The reinforcement learning model deep Q network (DQN) is used to process the network dynamic programming problem. The network dynamic programming includes state space definition, action space definition and reward mechanism design.

[0085] Further, regarding the definition of state space: in the scenario of network dynamic planning, the state space is composed of a plurality of said characteristic factors including said general characteristic factors (such as concurrent bandwidth adjustment factor, response delay adjustment factor and connection stability factor) and said business characteristic factors (such as traffic flow adjustment factor, medical data sensitivity factor and production load adjustment factor), and the specific state can be represented as a vector, which at least includes the current network status, network resource allocation and business demand information;

[0086] About the definition of action space: The action space defines various network resource allocation strategies that the system can adopt. That is, for each state, the reinforcement learning model deep Q network (DQN) needs to select a strategy in the network resource allocation strategy and assign an action;

[0087] Regarding the design of the reward mechanism: As the core part of network dynamic programming, the reward mechanism defines the feedback brought by each action. In the scenario of network dynamic programming, the reward mechanism is designed according to the following indicators:

[0088] ① If the selected action improves the service response speed, reduces latency, or reduces packet loss rate, a positive reward will be given;

[0089] ② If the selected action optimizes bandwidth utilization or reduces resource waste, a positive reward is given;

[0090] ③. In the transmission of sensitive data, ensuring the secure transmission of data can earn additional rewards;

[0091] The dynamic programming feature engineering model outputs an optimal network resource allocation strategy based on the input feature factors, and the network resource allocation strategy includes bandwidth allocation, routing path selection and transmission strategy adjustment;

[0092] Furthermore, the bandwidth allocation refers to dynamically adjusting the bandwidth allocation strategy between edge devices according to the bandwidth requirements of each service in the network, giving priority to increasing bandwidth allocation for key tasks;

[0093] The routing path selection refers to selecting a low-latency or high-security routing path according to the weight of the characteristic factor to ensure that data is transmitted in an optimal manner;

[0094] The transmission strategy refers to deciding whether to enable encrypted transmission, or to select other appropriate transmission methods to ensure data security and transmission efficiency.

[0095] See also Figure 2 In this embodiment, in S3, the experience replay mechanism means that after each action is executed, the system stores the tuple of state, action, reward and next state in the experience replay pool. This mechanism enables the dynamic programming feature engineering model to learn from past experiences, thereby breaking the correlation between data and improving learning efficiency;

[0096] Q-value updating means randomly extracting a batch of data from the experience replay pool for training each time, and updating the Q-value function through the Q-learning algorithm. DQN uses a deep neural network to approximate the Q-value function. The updating process includes:

[0097]

[0098] Among them: s is the current state, a is the current action, r is the immediate reward, s' is the next state, α is the learning rate, and γ is the discount factor;

[0099] The target network refers to the two neural networks used by the deep Q network of the reinforcement learning model. One is the main network, which is responsible for the current Q value prediction, and the other is the target network, which is used for regular updates to stabilize the training process. This mechanism helps to improve the convergence of the dynamic programming feature engineering model.

[0100] See also Figure 1-Figure 2 In this embodiment, S4 further includes:

[0101] S41, real-time monitoring: the system collects and obtains the characteristic factor data of the current network in real time through sensors or monitoring tools deployed on edge devices, and inputs it into the dynamic programming feature engineering model. The specific model parameters of the selected sensors or monitoring tools are not particularly limited in this embodiment, and the collection of characteristic factor data can be realized;

[0102] S42, real-time decision-making: After the dynamic programming feature engineering model is fully trained, the reinforcement learning model deep Q network (DQN) receives the current state information in real time and selects the optimal action based on the learned strategy. When the system detects changes in network status or fluctuations in business demand, the reinforcement learning model deep Q network quickly adjusts network resource allocation to adapt to the new environment, thereby maintaining high quality and stability of the service;

[0103] S43, self-update: when the network environment or business requirements are monitored to have major changes (such as bandwidth shortage or increased security requirements), the system will automatically re-plan the network resource allocation according to the new feature factor weights, and gradually optimize the dynamic programming feature engineering model during operation through the online learning algorithm adaptive gradient method (Adagrad);

[0104] S44, Feedback mechanism: The system compares the results of each network planning with the expected goals, and continuously adjusts the model parameters through the feedback mechanism to ensure that the allocation of network resources is always optimal.

[0105] Hereinafter, the present application will be described in detail in conjunction with the multiple specific business scenarios provided in the above embodiments:

[0106] Embodiment 1: Intelligent traffic scenario:

[0107] In the traffic monitoring system of smart cities, edge devices need to process a large amount of real-time video stream data from road cameras. The transmission of this data must ensure low latency and high bandwidth to achieve real-time and accurate monitoring.

[0108] Through the dynamic programming method of the present invention, the system can adjust the allocation of network resources in real time according to the changes in the current traffic flow. Specifically, when the traffic flow is large, the real-time demand for monitoring video data increases, and the system will automatically identify and extract traffic flow and video transmission delay as key characteristic factors. Based on these key characteristic factors, the system will prioritize the allocation of more network bandwidth to ensure smooth transmission and processing of monitoring data; conversely, when the traffic flow is small, the system will automatically reduce the allocation of bandwidth resources, thereby reducing network load and saving resources. This dynamic adjustment enables the traffic monitoring system to work efficiently under different traffic conditions and optimize resource usage.

[0109] Example 2: Smart medical scenario:

[0110] In smart medical scenarios, edge devices are responsible for processing and transmitting large amounts of patient health data, which are both sensitive and have high requirements for real-time transmission.

[0111] Through the dynamic programming method of the present invention, the system can dynamically allocate network resources based on extracted characteristic factors (such as data sensitivity, delay requirements, etc.) in different medical business scenarios. Specifically, when the system detects that the patient data to be transmitted is sensitive information, it automatically enables the encrypted transmission channel to ensure data privacy and security. At the same time, in emergency medical scenarios, such as remote surgery or real-time monitoring, the system will give priority to allocating high-bandwidth and low-latency network resources to ensure the timely transmission of critical medical information. Conversely, in the case of ordinary examination data or non-sensitive scenarios, the system reduces the occupancy of network resources and optimizes bandwidth allocation, so that the medical system can still achieve efficient resource management and transmission guarantee while ensuring patient privacy and safety.

[0112] Example 3: Smart factory scenario:

[0113] During the production process of smart factories, edge devices need to transmit a large amount of production status data in real time, and as the factory operation conditions change, the system needs to dynamically adjust network resources.

[0114] Through the dynamic programming method of the present invention, the system can extract characteristic factors such as equipment load and data sensitivity according to different scenarios of production load. Specifically, when the factory is in a high-load production stage, the demand for data communication between devices increases. The system will identify the priority of low latency and high bandwidth, and allocate more network resources to them to ensure the stable operation of the production control system. At the same time, during the production process, if the data processed by the equipment involves sensitive information, the system will automatically select an encrypted channel for transmission to ensure the security of the data. Conversely, in the case of non-high-load or non-sensitive data transmission, the system reduces bandwidth allocation to reduce energy consumption, thereby improving the overall communication efficiency and security of the factory.

[0115] An embodiment of the present invention further provides a computer-readable storage medium storing a computer program. When the computer program is executed by a processor, the processor executes the steps of the above method.

[0116] See also Figure 3 An embodiment of the present invention further provides a computer device, including a memory and a processor, wherein the memory stores a computer program, and when the computer program is executed by the processor, the processor executes the steps of the above method.

[0117] The embodiment of the present invention further provides a computer program product comprising instructions, which, when executed on a computer, enables the computer to execute the steps of the above method.

[0118] It is understandable that the system, device and storage medium provided in the embodiments of the present invention correspond to the method provided in the embodiments of the present invention, and the explanation, examples and beneficial effects of the relevant contents can refer to the corresponding parts in the above methods.

[0119] In summary, the present invention proposes a network dynamic planning method based on characteristic factors, which is specifically used for network operation and management of edge devices in IoT scenarios. By extracting and analyzing characteristic factors in different business scenarios, dynamic adjustment and optimization of network resources are achieved. The present invention aims to solve the problem of lack of perception of specific business scenario requirements and personalized scheduling in existing network planning methods, so that edge devices can flexibly adjust network strategies according to scenario changes, so as to improve the utilization efficiency of network resources, enhance the security of data transmission and reduce communication delays.

[0120] Specifically, this method uses characteristic factors for dynamic planning, and the system can automatically identify business needs in different scenarios, such as enabling encrypted channels in data-sensitive scenarios, automatically adjusting bandwidth in high-concurrency scenarios, and prioritizing critical data transmission in low-latency scenarios. This method not only improves the adaptability of network planning, but also runs efficiently on edge devices with limited resources, meeting the dynamic needs of complex business scenarios.

[0121] It should be understood that the examples and implementation modes described herein are for illustrative purposes only and are not intended to limit the present invention. Those skilled in the art may make various modifications or changes based on the examples and implementation modes. Any modifications, equivalent substitutions, improvements, etc. made within the spirit and principles of the present invention should be included in the scope of protection of the present invention.

Claims

1. A network dynamic planning method for edge IoT devices based on characteristic factors, characterized in that: The following steps are involved: S1. Extraction and weighting of characteristic factors: Extract general characteristic factors and service characteristic factors, and assign weights to the extracted characteristic factors to assist in scheduling network resources. S2. Constructing a dynamic programming feature engineering model: constructing the dynamic programming feature engineering model based on the feature factors obtained in S1, wherein the dynamic programming feature engineering model uses a reinforcement learning model deep Q network to process network dynamic programming problems, adjust and output the optimal network resource allocation strategy, and dynamically adapt network resources to business needs; S3, training a dynamic programming feature engineering model: based on the dynamic programming feature engineering model obtained in S2, at least one or more of the experience replay mechanism, Q value update or target network is used to perform model training; S4. Optimize the dynamic programming feature engineering model: update the feature factor data input into the dynamic programming feature engineering model in real time. The dynamic programming feature engineering model re-plans network resource allocation when the network status changes or the business demand fluctuates, and gradually optimizes and updates during operation.

2. The network dynamic planning method based on characteristic factors of edge IoT devices as claimed in claim 1 is characterized in that: In S1, the general characteristic factors are used to characterize general service indicators in general service scenarios, including at least a concurrent bandwidth adjustment factor, a response delay adjustment factor, and a connection stability factor; The concurrent bandwidth adjustment factor is composed of the number of concurrent business services, the current bandwidth occupancy rate and the service request rate. When the system detects that the number of concurrent business services has increased, combined with the current bandwidth occupancy rate and the service request rate, the system evaluates and allocates more bandwidth resources to high-concurrency services in real time to ensure smooth data transmission and avoid congestion and delay. The response delay adjustment factor is combined with the business service response time, the current network delay and the current load. When the business service response time exceeds the set threshold, the system extracts the characteristics of the current network delay and the current load, determines whether it is necessary to adjust the routing path or increase bandwidth resources, optimizes the network configuration, and reduces the delay of service response; The connection stability factor is used to monitor the stability of the network and the reliability of the service connection. It is composed of the business service connection loss rate, data packet loss rate and network transmission error rate. When the business service connection loss rate or data packet loss rate is frequent, the system detects and adjusts the network's transmission strategy to ensure the stable operation of key businesses.

3. The network dynamic planning method based on characteristic factors of edge IoT devices as claimed in claim 1 is characterized in that: In S1, the business-specific diagnostic factors are used to characterize specific business indicators in specific business scenarios, including at least a data sensitivity factor, a real-time demand factor, and a concurrent number factor of standby communication; The data sensitivity factor is used to determine whether the data needs to be transmitted through an encrypted channel; The real-time requirement factor is used to control the low-delay requirement of the signal to ensure the instant response of the business operation; The concurrency factor of the communication between the backup devices is used to determine the dynamic allocation demand of the bandwidth.

4. The network dynamic planning method based on characteristic factors of edge IoT devices as claimed in claim 3 is characterized in that: The specific business scenarios represented by the business-specific diagnosis factors include at least smart traffic scenarios, smart medical scenarios and smart factory scenarios, and the corresponding characteristic factors are traffic flow adjustment factors, medical data sensitivity factors and production load adjustment factors respectively; The traffic flow adjustment factor is composed of real-time traffic flow, vehicle density and camera video data frame rate. In the intelligent traffic scenario, when the traffic flow increases, the system extracts the factor to dynamically adjust the bandwidth resources. The vehicle density and camera video data frame rate jointly determine the demand for bandwidth. The system ensures the real-time and clarity of the video stream by increasing bandwidth allocation, avoiding delays or freezes in the monitoring screen due to insufficient bandwidth. The medical data sensitivity factor is composed of the patient data sensitivity level, real-time monitoring requirements and data transmission index requirements. In the smart medical scenario, the system extracts the factor to perform sensitivity assessment on the patient data. When the system detects that the patient data sensitivity level is highly sensitive data, the medical data sensitivity factor triggers the encryption transmission strategy and, based on the real-time monitoring requirements, prioritizes the allocation of high-bandwidth and low-latency network resources to ensure that the data transmission index requirements are met. The production load adjustment factor is combined with the production line equipment load rate, real-time production data volume and operation task priority. In the smart factory scenario, when the production line equipment load rate increases, the system extracts the factor to evaluate the real-time load status and real-time production data volume of the production line equipment, ensuring that high-level critical tasks in the operation task priority meet the network resource requirements of high bandwidth and low latency.

5. The network dynamic planning method based on characteristic factors of edge IoT devices as claimed in claim 1 is characterized in that: In S1, the weight ratio of the characteristic factors is optimized by a weighted linear model of a multi-objective optimization algorithm based on historical data.

6. The network dynamic planning method based on characteristic factors of edge IoT devices as claimed in claim 1, characterized in that: The S2 further comprises: S21, data cleaning: before constructing the dynamic programming feature engineering model, preprocess the extracted feature factors to ensure the accuracy and consistency of the data; Data cleaning operations at least include missing data processing, abnormal data filtering and duplicate data removal, among which missing data processing is to fill the missing data with the mean or median, and abnormal data filtering and duplicate data removal are to identify and filter outlier data through z-score or clustering-based methods; S22, model construction: The core of the dynamic programming feature engineering model is network dynamic programming based on the feature factors, and the reinforcement learning model deep Q network is used to process the network dynamic programming problem. The network dynamic programming includes state space definition, action space definition and reward mechanism design; The dynamic programming feature engineering model outputs an optimal network resource allocation strategy based on the input feature factors, and the network resource allocation strategy includes bandwidth allocation, routing path selection and transmission strategy adjustment.

7. The network dynamic planning method based on characteristic factors of edge IoT devices as claimed in claim 6, characterized in that: In the S22, in the scenario of network dynamic planning, the state space is composed of a plurality of the characteristic factors including the general characteristic factors and the service characteristic factors, and the specific state can be represented as a vector, which at least includes the current network status, network resource allocation status and service demand information; The action space defines the various network resource allocation strategies that the system can adopt, that is, for each state, the reinforcement learning model deep Q network needs to select a strategy in the network resource allocation strategy and assign an action; As the core part of network dynamic programming, the reward mechanism defines the feedback brought by each action. In the scenario of network dynamic programming, the reward mechanism is designed according to the following indicators: ① If the selected action improves the service response speed, reduces latency, or reduces packet loss rate, a positive reward will be given; ② If the selected action optimizes bandwidth utilization or reduces resource waste, a positive reward is given; ③. In the transmission of sensitive data, additional rewards can be obtained by ensuring the secure transmission of data.

8. The network dynamic planning method based on characteristic factors of edge IoT devices as claimed in claim 6, characterized in that: In S22, the bandwidth allocation refers to dynamically adjusting the bandwidth allocation strategy between edge devices according to the bandwidth requirements of each service in the network, and giving priority to increasing bandwidth allocation for key tasks; The routing path selection refers to selecting a low-latency or high-security routing path according to the weight of the characteristic factor to ensure that data is transmitted in an optimal manner; The transmission strategy refers to deciding whether to enable encrypted transmission, or to select other appropriate transmission methods to ensure data security and transmission efficiency.

9. The network dynamic planning method based on characteristic factors of edge IoT devices as claimed in claim 7, characterized in that: In S3, the experience replay mechanism means that after each action is executed, the system stores the tuple of state, action, reward and next state in the experience replay pool; Q-value updating means randomly extracting a batch of data from the experience replay pool for training each time, and updating the Q-value function through the Q-learning algorithm; The target network refers to the two neural networks used by the deep Q network of the reinforcement learning model. One is the main network, which is responsible for the current Q value prediction, and the other is the target network, which is used for regular updates to stabilize the training process.

10. The network dynamic planning method based on characteristic factors of edge IoT devices according to claim 1, characterized in that: The S4 further comprises: S41, real-time monitoring: the system collects and obtains the characteristic factor data of the current network in real time through sensors or monitoring tools deployed on edge devices, and inputs the data into the dynamic programming feature engineering model; S42, real-time decision-making: After the dynamic programming feature engineering model is fully trained, the reinforcement learning model deep Q network receives the current state information in real time and selects the optimal action based on the learned strategy. When the system detects changes in network state or fluctuations in business demand, the reinforcement learning model deep Q network quickly adjusts network resource allocation to adapt to the new environment; S43, self-update: when a significant change in the network environment or business demand is detected, the system will automatically re-plan the network resource allocation according to the new feature factor weights, and gradually optimize the dynamic programming feature engineering model during operation through the online learning algorithm adaptive gradient method; S44, Feedback mechanism: The system compares the results of each network planning with the expected goals, and continuously adjusts the model parameters through the feedback mechanism to ensure that the allocation of network resources is always optimal.

Images