Cloud system capable of quickly accessing edge device based on Internet of Things

Abstract

The invention relates to the technical field of Internet of Things, and discloses an Internet of Things-based cloud system capable of quickly accessing an edge device, and the system comprises an edge device access module which is used for receiving an access request of the edge device and completing device authentication and communication protocol adaptation; the cloud resource monitoring module is used for monitoring resource use states of a cloud in real time, including bandwidth, computing power and use conditions of storage resources; the cloud optimization scheduling module is used for dynamically allocating cloud resources based on the resource requirements and priorities of the edge devices; the data feedback and adjustment module is used for feeding back the resource allocation result to the edge device and dynamically adjusting the transmission rate of the edge device; and the data management and security module is used for storing historical data and distribution schemes of the edge devices. Through the modular design, the dynamic optimization scheduling and the differential privacy technology, the rapid access of the edge equipment, the efficient resource allocation and the comprehensive guarantee of the data security and privacy are realized.

Description

Technical Field

[0001] The present invention relates to the technical field of Internet of Things, and in particular to a cloud system based on the Internet of Things that can quickly access edge devices. Background Art

[0002] With the widespread application of IoT technology, edge devices are playing an increasingly important role in the fields of industry, agriculture, energy, etc. For example, the edge controller in the distributed energy storage system needs to transmit device data to the cloud for real-time monitoring, resource allocation, and data processing. In order to adapt to the trend of device diversification and complex requirements, the cloud system needs to have efficient device access capabilities and be able to dynamically manage resource allocation to meet changing business needs.

[0003] Existing technologies have proposed some solutions for edge device access and resource management, which have played a good role in specific scenarios. For example, some cloud systems have achieved access and data transmission for some devices through customized protocol adapter modules; other resource management technologies use static allocation strategies to provide stable performance in simple application scenarios. In addition, some security mechanisms, such as symmetric encryption and key-based authentication technology, can improve data security to a certain extent. These technologies provide basic support for the initial development of the Internet of Things.

[0004] However, the existing technology still has obvious deficiencies in many aspects, which limits its application in complex IoT scenarios. First, in the process of device access, the existing solutions lack flexible standardized design. When facing heterogeneous devices, it is necessary to independently develop adaptation modules, resulting in low development efficiency and poor scalability. Secondly, the static resource allocation method cannot adapt to the dynamic changes in device demand. It may cause insufficient resources when the load suddenly increases, and easily lead to resource waste when the load decreases. In addition, the data security design in the existing technology is mostly concentrated on simple encryption schemes in a single scenario, which is difficult to simultaneously meet the privacy protection and data integrity requirements when multiple devices are connected concurrently. Finally, the insufficient prediction ability of device demand is a major bottleneck of the existing technology. Traditional static analysis based on historical data cannot accurately capture the nonlinear changes in device demand, which directly affects the efficiency and accuracy of resource scheduling. Summary of the invention

[0005] In view of the shortcomings of the prior art, the present invention provides a cloud system based on the Internet of Things that can quickly access edge devices, solving the problems in the prior art of complex edge device access process, poor scalability, static resource allocation method unable to cope with dynamic demand changes, and insufficient data security and privacy protection capabilities.

[0006] To achieve the above objectives, the present invention is implemented through the following technical solutions: A cloud system based on the Internet of Things that can quickly access edge devices, including: The edge device access module is used to receive access requests from edge devices and complete device authentication and communication protocol adaptation; The cloud resource monitoring module is used to monitor the resource usage status of the cloud in real time, including the usage of bandwidth, computing power and storage resources; Cloud optimization scheduling module, which is used to dynamically allocate cloud resources based on the resource requirements and priorities of edge devices; The data feedback and adjustment module is used to feed back the resource allocation results to the edge device and dynamically adjust the transmission rate of the edge device; the data management and security module is used to store the historical data and allocation plan of the edge device and to provide security protection for data transmission and storage.

[0007] Preferably, the edge device access module includes: The device authentication submodule is used to complete device authentication based on the unique identifier of the edge device to ensure the legitimacy of device access; the communication protocol adaptation submodule is used to support different communication protocols and convert the resource demand information sent by the edge device into a data format that can be processed by the cloud; The demand upload submodule is used to upload the resource demand information and priority weights of edge devices to the cloud optimization scheduling module.

[0008] Preferably, the cloud resource monitoring module includes: The resource status collection submodule is used to collect the bandwidth, computing power and storage resource usage of the cloud system in real time; the device status monitoring submodule is used to monitor the current operating status of the edge device, including resource allocation and transmission status; the historical data recording submodule is used to record the historical resource requirements and allocation plans of the edge device to provide a reference for subsequent optimization.

[0009] Preferably, the cloud optimization scheduling module includes: The state prediction submodule is used to predict future resource demand based on the historical demand data and current resource status of edge devices; the resource allocation submodule is used to dynamically calculate the resource allocation plan according to the priority weight and demand of edge devices; The iterative scheduling submodule is used to dynamically adjust the allocation parameters according to the allocation results to achieve global balance of resources.

[0010] Preferably, the resource demand prediction step of the state prediction submodule includes: Collect historical resource demand data of edge devices; Generate future resource requirements based on the current state of edge devices and random disturbance factors; The prediction results are provided to the resource allocation submodule.

[0011] Preferably, the resource allocation step of the resource allocation submodule includes: Calculate the resource allocation for each device based on the resource requirements, priority weights, and resource status of the edge device; Ensure that all resource allocations meet the total resource constraints of the cloud system; Optimize resource allocation to ensure that the needs of high-priority devices are met first while ensuring fairness for other devices.

[0012] Preferably, the data feedback and adjustment module includes: The resource allocation feedback submodule is used to feed back the resource allocation results generated by the cloud optimization scheduling module to the edge device; The transmission rate adjustment submodule is used to adjust the transmission rate of edge devices to balance the network load when resources are insufficient; The dynamic compensation submodule is used to adjust the allocation scheme to meet the new demand when the demand of edge devices changes.

[0013] Preferably, the dynamic resource compensation step of the dynamic compensation submodule includes: Detect the current resource usage of edge devices; Determine whether device demand has increased or network load has decreased; Recalculate the allocation plan based on the remaining resources in the cloud and adjust the device allocation resources.

[0014] Preferably, the data management and security module includes: The data storage submodule is used to store the resource requirements, allocation schemes and historical data of edge devices; The data encryption submodule is used to encrypt data during data transmission to prevent data leakage; The data authentication submodule is used to verify the legitimacy of edge devices and ensure that the data source is credible.

[0015] Preferably, the privacy protection function of the data management and security module includes the following steps: The sensitive data uploaded by edge devices is processed in a graded manner, and sensitive information is stored separately from general data; Perform differential privacy processing on sensitive data and add random noise to blur the processed results; The processed data is used for resource allocation calculations to ensure the privacy and security of the device.

[0016] The present invention provides a cloud system based on the Internet of Things that can quickly access edge devices. It has the following beneficial effects: 1. The present invention adopts modular design technology to organically integrate edge device access, cloud resource monitoring, optimized scheduling, feedback regulation, data management and security functions to form an overall architecture for collaborative operation. This design makes device access faster, improves resource utilization, and makes data management more reliable. Compared with traditional decentralized design, it solves the problems of slow response and poor scalability.

[0017] 2. The present invention uses the dynamic Lagrangian optimization technology in the cloud optimization scheduling module to dynamically allocate resources according to device priority and demand. It can make resource allocation more efficient and achieve global balance. Compared with the existing static allocation method, it solves the problem of insufficient resource utilization and inflexible adjustment.

[0018] 3. The present invention significantly enhances data security and privacy protection capabilities by introducing differential privacy and distributed storage technology. Random noise is added when the device uploads data, and distributed storage is used to avoid single point failures. This method effectively prevents the leakage of sensitive information, and also makes data storage more stable and access more efficient, which has obvious advantages over existing centralized storage methods. BRIEF DESCRIPTION OF THE DRAWINGS

[0019] Figure 1 is a system structure diagram of the present invention; Figure 2 This is a module architecture diagram of the edge device access module of the present invention; Figure 3 This is a module architecture diagram of the cloud resource monitoring module of the present invention; Figure 4 This is a module architecture diagram of the cloud-based optimization scheduling module of the present invention; Figure 5 This is a module architecture diagram of the data feedback and adjustment module of the present invention; Figure 6 This is a module architecture diagram of the data management and security module of the present invention. DETAILED DESCRIPTION

[0020] The following will be combined with the drawings in the specification of the present invention to clearly and completely describe the technical solutions in the embodiments of the present invention. Obviously, the described embodiments are only part of the embodiments of the present invention, not all of the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by ordinary technicians in this field without creative work are within the scope of protection of the present invention.

[0021] Please see attached Figure 1 -Attached Figure 6 , an embodiment of the present invention provides a cloud system based on the Internet of Things that can quickly access edge devices, including: The edge device access module is used to receive access requests from edge devices and complete device authentication and communication protocol adaptation; The edge device access module ensures that the device can interact with the cloud for resource demand information under legal and secure conditions through device authentication, communication protocol adaptation, and demand upload functions. Generally, this module receives access requests from edge devices, parses their resource demand data into a standardized format, and uploads it to the cloud optimization and scheduling module. As an option, the edge device access module collaborates with the cloud resource monitoring module to ensure the legitimacy of the access device and provide accurate data support for subsequent resource allocation.

[0022] In this embodiment, the structure and implementation of the edge device access module are as follows: In the initial stage of edge device access, the device authentication submodule is mainly responsible for verifying the legitimacy of the device and preventing illegal devices from accessing the cloud system. Specifically, this submodule uses the unique identification information of each device (such as device ID, digital certificate or pre-shared key) for identity authentication.

[0023] In a possible implementation, the device authentication submodule adopts an authentication mechanism based on symmetric encryption or asymmetric encryption. For example, the device sends an encrypted information request containing its unique identifier to the cloud, and the cloud confirms its legitimacy by decrypting it.

[0024] Generally, the device authentication submodule also supports dynamic key negotiation to enhance the security of the authentication process. In some embodiments, after authentication, a session key is generated for the device, which is used for encryption protection of subsequent data transmission.

[0025] After the device authentication is completed, the communication protocol adapter module is responsible for adapting the communication protocol used by the device and converting the data uploaded by the device into a unified format that can be recognized by the cloud system. Specifically, the communication protocol adapter module can support common IoT protocols, such as MQTT and TCP / IP protocols.

[0026] As an option, the communication protocol adapter submodule can automatically select an adapted protocol stack according to the device type. In a possible implementation, for a device that communicates based on the MQTT protocol, the submodule extracts resource demand data by parsing the topic and message body sent by the device.

[0027] In another case, for a device based on the TCP / IP protocol, the submodule directly receives data packets and extracts the resource requirement information of the device by parsing the content of the data frame.

[0028] To ensure that the data can be correctly processed by the cloud system in the future, this module will convert the extracted data into a standardized JSON format. The fields in the JSON data format include device identification, demand timestamp, resource demand type (such as bandwidth, computing power, storage capacity) and priority weight.

[0029] After the communication protocol adaptation is completed, the demand upload submodule will upload the parsed resource demand data to the cloud optimization scheduling module. Specifically, this module determines the priority weight of the device according to the following formula: Where: w i : The priority weight of device i; P i : The priority parameter of device i, which is determined by the service type or urgency of the device; N: The total number of edge devices currently connected; The sum of the priority parameters of all connected devices, which is used to normalize the priority weight of each device, Ensure that the sum of the weights of all devices is equal to 1; j: represents the device index currently being calculated, j ranges from 1 to N; N: the total number of edge devices currently connected; Generally, the demand upload submodule packages the resource demand and priority weight together and uploads it to the cloud system through a secure channel. Specifically, the data package includes the following: Unique device identifier, used to distinguish different devices; Timestamp, used to mark the request time of resource requirements; Resource Requirement Vector Priority weight w i .

[0030] The above resource demand vector is defined as follows: The bandwidth requirement of device i at time t (in Mbps); The computing power required by device i at time t (unit: number of CPU cores); Storage capacity requirement of device i at time t (in GB).

[0031] In a possible implementation, the demand upload submodule compresses the uploaded data to reduce network bandwidth usage. As an option, the module also supports a multi-threaded upload mechanism that can handle resource requirements of multiple devices at the same time.

[0032] As a further technical extension, the edge device access module also supports dynamic resource estimation. Specifically, in the communication protocol adapter submodule, a real-time data monitoring function can be added to analyze the historical demand trend of the device and generate future demand forecasts based on the current usage. For example, through the following linear prediction model: in: The predicted demand of equipment i at time t+1; The actual demand of device i at time t; k: adjustment factor used to control the sensitivity of the prediction, usually in the range of 0<k≤1.

[0033] Through the above method, the system can predict the resource usage of equipment before the demand is submitted, thereby providing more accurate decision support for the cloud optimization scheduling module.

[0034] In this embodiment, the edge device access module ensures that the edge device can quickly and securely access the cloud system through device authentication, communication protocol adaptation and demand upload functions.

[0035] The cloud resource monitoring module is used to monitor the resource usage status of the cloud in real time, including the usage of bandwidth, computing power and storage resources; In the technical solution of the present invention, the cloud resource monitoring module is an important component for realizing resource management and dynamic scheduling. This module is mainly responsible for real-time collection and monitoring of the resource usage of the cloud system, including key resource information such as bandwidth, computing power and storage capacity, while tracking the operating status of the connected devices, and providing accurate input data for the cloud optimization scheduling module. Generally, this module collaborates with the edge device access module, can dynamically update the usage of system resources, and provide support for subsequent resource allocation strategies through historical data records. As an option, the module can also perform load optimization processing on the access device based on the collected resource status information.

[0036] In this embodiment, the specific implementation of the cloud resource monitoring module is as follows: The resource status collection submodule is the basic functional part of the cloud resource monitoring module, which is used to obtain the system resource usage in real time.

[0037] Specifically, this submodule collects the current usage and remaining available amount of bandwidth, computing power and storage capacity in the cloud system through a multi-threaded monitoring mechanism. As an option, the collection frequency can be dynamically adjusted according to changes in system load to reduce resource consumption.

[0038] In a possible implementation, the submodule sets independent collection rules for each resource. For example: Bandwidth resources measure the total bandwidth usage of current uploads and downloads in real time through traffic monitoring tools.

[0039] The total usage of computing resources is calculated by counting the occupancy rate of each CPU core.

[0040] Storage resources determine the remaining storage space by querying the capacity usage of the storage pool.

[0041] In addition, in order to ensure the accuracy of resource status collection, this submodule will also pre-process the collected data. For example, for abnormal fluctuations caused by burst traffic, the data can be smoothed by weighted moving average method, the formula is as follows: R avg (t) = αR(t) + (1-α)R avg (t-1) Where: R avg (t): smoothed resource utilization rate at time t; R(t): actual resource utilization rate at time t; α: smoothing coefficient, the value range is 0<α≤1.

[0042] The device status monitoring submodule is responsible for tracking the operation of connected devices, including their resource allocation and transmission status.

[0043] Generally, this submodule will periodically synchronize the status with the connected device. The synchronization content includes the current resource consumption of the device, the progress of task completion, and whether there are transmission anomalies. In one possible implementation, this submodule implements status monitoring through a heartbeat packet mechanism. For example, the device periodically sends a heartbeat signal to the cloud, which contains the device's resource usage statistics. The cloud obtains the real-time status of the device by parsing the signal content.

[0044] Specifically, for each device i, the device status monitoring submodule records the following key data: The resources currently allocated to device i; ΔR i (t): the change in resource usage of device i; Whether the transmission rate meets expectations, if not, an alarm mechanism is triggered.

[0045] As an option, the equipment status monitoring submodule can also combine historical data to predict the trend of the equipment's operating status. For example, the resource usage growth of the equipment can be predicted through time series analysis methods, thereby warning of possible resource shortages.

[0046] The historical data recording submodule is used to store the resource usage history data of the cloud system and the operation history data of the equipment, providing a reference for subsequent optimization and scheduling.

[0047] Specifically, this submodule logs the following: The usage history of the total system resources, including the peak value, valley value, and daily average value of each resource.

[0048] Resource allocation and usage history for each device, including demand data and actual allocation results when the device was connected.

[0049] In one possible implementation, the historical data recording submodule implements data storage through a distributed database to support high-concurrency data query and write operations. For example, for bandwidth resources, hourly usage statistics can be recorded and usage trends within a specified time period can be quickly returned when querying.

[0050] In addition, this submodule also supports aggregate analysis of historical data. For example, it can calculate the average resource requirements of each device type to assist the optimization scheduling module in formulating a more reasonable resource allocation strategy.

[0051] As a further optimization solution, the cloud resource monitoring module can also integrate resource prediction functions. Based on the device status monitoring submodule, a prediction algorithm based on machine learning is added to dynamically adjust the system's resource acquisition and allocation strategies. For example, resource demand trends can be predicted by training regression models, and allocation parameters can be adjusted in advance based on the current resource status.

[0052] In this embodiment, the cloud resource monitoring module provides basic support for the system's resource management and dynamic optimization through resource status collection, device status monitoring and historical data recording functions.

[0053] The cloud optimization scheduling module is used to dynamically allocate cloud resources based on the resource requirements and priorities of edge devices. The cloud optimization scheduling module dynamically calculates the optimal resource allocation plan based on the resource status information provided by the cloud resource monitoring module and the device requirement information uploaded by the edge device access module, combined with priority weights and system constraints. In general, this module achieves reasonable allocation of global resources through state prediction, resource allocation calculation and iterative scheduling. As an option, this module can also readjust the resource allocation plan according to the real-time changes in the operating status of the device to ensure the maximum utilization of system resources and the dynamic balance of device requirements.

[0054] In this embodiment, the implementation of the cloud-based optimization scheduling module includes a state prediction submodule, a resource allocation submodule and an iterative scheduling submodule.

[0055] The main function of the state prediction submodule is to predict future resource demand based on the historical demand data and current resource status of edge devices. Generally, this submodule provides a reference for resource allocation in advance by analyzing the changing trend of device resource demand.

[0056] Specifically, this submodule collects historical data on resource requirements of each device at different points in time and analyzes it in combination with the current resource usage of the system; As an option, this submodule can also adjust the weight of the prediction results according to the task type and priority parameters of the device. For example, for high-priority devices, the prediction results can be appropriately increased to ensure the reliability of resource allocation; The task of the resource allocation submodule is to calculate the optimal resource allocation plan based on the prediction results provided by the status prediction submodule and the resource status information provided by the cloud resource monitoring module.

[0057] In general, the resource allocation submodule allocates resources based on the priority weight of the device and the total resource constraint. The goal of the allocation calculation is to maximize the system utility while meeting the minimum resource requirements of all devices. The utility function is defined as: Where: U: total utility of the system; w i : The priority weight of device i; The resources actually allocated to device i; Resource requirements of device i.

[0058] The resource allocation plan must meet the following constraints: Total resource limit: The sum of allocated resources of all devices must not exceed the total system resources; Minimum requirement guarantee: The allocated resources for each device must not be lower than its minimum requirement; Fairness constraint: The allocation ratio of high-priority devices should be given priority, but low-priority devices cannot be completely excluded.

[0059] Specifically, the resource allocation submodule uses an iterative optimization algorithm to solve the allocation results. For example, in the initial stage, the system allocates the minimum required amount of resources to each device. Subsequently, the system gradually adjusts the allocation results according to the priority weights and remaining resources until the allocation results converge.

[0060] The iterative scheduling submodule is responsible for readjusting the resource allocation plan according to the changes in the device operating status after the resource allocation calculation is completed. Generally, this submodule ensures that the allocation results always meet the global needs of the system by dynamically adjusting the optimization parameters.

[0061] As an implementation method, the iterative scheduling submodule adopts a dynamic adjustment method based on Lagrangian optimization. Specifically, this method dynamically updates the allocation parameters through the following formula: Where: λ: Lagrange multiplier, used to balance the total resource constraints; η: learning rate, controlling the step size of parameter adjustment; Rtotal : The total amount of resources in the cloud system; The total amount of allocated resources for all connected devices; N: the number of edge devices currently connected.

[0062] As an option, the iterative scheduling submodule can also make fine-grained adjustments to the allocation results based on the real-time transmission rate and task progress of the device. For example, for devices with lower-than-expected transmission rates, their bandwidth allocation can be reduced; for devices with delayed task progress, their computing resource allocation can be increased.

[0063] In order to further improve allocation efficiency, the cloud optimization scheduling module can introduce machine learning technology to learn the relationship between historical allocation results and actual resource usage. For example, by training a regression model to predict the optimal adjustment direction of the allocation results, the number of iterations can be reduced. In addition, in the resource allocation calculation, the network topology information of the device can be combined to give priority to allocating bandwidth to devices with longer transmission distances, thereby reducing the total transmission delay.

[0064] In this embodiment, the cloud optimization scheduling module realizes dynamic allocation and global optimization of resources through the collaboration of the state prediction submodule, the resource allocation submodule and the iterative scheduling submodule.

[0065] Data feedback and adjustment module, used to feed back resource allocation results to edge devices and dynamically adjust the transmission rate of edge devices; The data feedback and adjustment module is connected between the cloud optimization scheduling module and the edge device. Its main task is to feed back the optimization results to the device and dynamically adjust and compensate according to the actual operation of the device. Generally, this module includes three functions: resource allocation feedback, transmission rate adjustment, and dynamic compensation. As an option, this module can also combine the real-time status information of the device and the remaining resources of the system to make fine-grained adjustments to resource allocation to ensure efficient matching of the actual needs of the device with the allocated resources.

[0066] In this embodiment, the specific implementation of the data feedback and adjustment module includes the following parts: The resource allocation feedback submodule is used to accurately transmit the resource allocation plan generated by the cloud optimization scheduling module to the edge device. Specifically, this submodule sends the resource allocation results to each device in the form of structured data through the IoT message middleware or direct communication.

[0067] In a possible implementation, the submodule packages the allocation results into a data packet in the following format: Device ID i : Uniquely identifies device i; Allocate resources The resources actually obtained by device i, including bandwidth, computing power, and storage capacity; Timestamp T: records the generation time of the resource allocation result, which is used to ensure the real-time nature of the allocation result.

[0068] The following is an example of a data packet: Specifically, the data packet is transmitted to the device through a secure communication protocol, and the device understands the resource allocation result by parsing the data packet content. In general, this submodule also verifies the allocation result to ensure the consistency of the allocation plan with the device requirements.

[0069] The transmission rate adjustment submodule is mainly used to adjust the data upload rate of the device when the device resource demand does not match the actual allocation result. Generally, when system resources are tight or the device does not get the expected resource allocation, this submodule will limit the data transmission rate of the device to reduce resource usage.

[0070] In a possible implementation, the submodule dynamically adjusts the data upload rate V of device i according to the following rules: i : Where: V i : Actual upload rate of device i; The actual amount of resource allocation obtained by device i; Resource requirements of device i; V i max : Maximum upload rate of device i.

[0071] Specifically, when When the device uploads data, the upload rate will be appropriately reduced to avoid occupying too much system bandwidth resources.

[0072] As an option, the submodule can also dynamically monitor the actual upload rate of the device, compare it with the expected rate, and trigger an alarm mechanism when the deviation exceeds a threshold. For example, when the actual upload rate of the device is less than 80% of the expected value, the submodule will send an alarm signal to the cloud.

[0073] The dynamic compensation submodule is used to adjust the allocation results in real time when the device demand changes or the resource allocation is not fully satisfied. Generally, when the device demand increases or the system resource situation improves, this submodule will reallocate the remaining resources and provide additional resource compensation to the device.

[0074] Specifically, the workflow of this submodule includes the following steps: Monitor the actual operating status of the equipment, including resource utilization and task completion progress; Determine whether equipment demand has increased or system load has decreased; The compensation plan is calculated based on the remaining resources in the cloud, and the additional allocated resources are fed back to the device.

[0075] In a possible implementation, the submodule calculates the compensation resource amount ΔR of device i according to the following formula: i : Where: ΔR i : Compensation resource amount of device i; Resource requirements of device i; The resources allocated to device i; R rem : Remaining available resources in the cloud.

[0076] As an option, this submodule can also combine the priority weight of the device to give priority to meeting the compensation needs of high-priority devices. For example, for the priority weight w i For equipment with a RMS of >0.8, the amount of compensation resources can be increased by 20% in proportion.

[0077] To further optimize the compensation process, the module can adjust the compensation strategy based on the device's real-time task type. For example, for tasks with high real-time requirements, the dynamic compensation submodule can prioritize bandwidth resources; for storage tasks, it prioritizes storage capacity. In addition, after the device task is completed, the module can recycle unused resources and allocate them to other devices.

[0078] In this embodiment, the data feedback and adjustment module ensures that the resource allocation result can be accurately executed and realizes the dynamic matching between device demand and resource allocation through resource allocation feedback, transmission rate adjustment and dynamic compensation functions.

[0079] Data management and security module, used to store historical data and allocation plans of edge devices, and to provide security protection for data transmission and storage; The data management and security module works closely with the data feedback and adjustment module. On the one hand, it stores the demand data uploaded by the device and the resource allocation plan generated by the system. On the other hand, it ensures the security and reliability of data transmission through multi-layer encryption and authentication mechanisms. Generally, this module includes a data storage submodule, a data encryption and authentication submodule, and a privacy protection submodule. As an option, this module can also combine distributed storage technology to optimize data access efficiency, and use differential privacy technology to protect sensitive information of the device while ensuring the effectiveness of data analysis.

[0080] The data storage submodule is responsible for recording key data during the system operation, including the resource demand information uploaded by the edge devices and the resource allocation plan generated by the system.

[0081] Specifically, this submodule creates a separate data storage area for each device to store the following data: Resource Requirement Vector represents the resource requirements of device i at time t, including bandwidth, computing power, and storage capacity.

[0082] Allocation results Record the actual resource allocation amount obtained by device i at time t.

[0083] Historical records: including the operating status of the equipment, resource demand change trends and task completion status.

[0084] In one possible implementation, the submodule uses distributed storage technology to store data on multiple nodes. Generally, this approach can improve data access speed and reliability. For example, for high-frequency resource allocation data, the system will replicate data between multiple nodes to reduce the latency of single-point access.

[0085] As an option, the submodule can also automatically classify storage according to data type. For example, data of high-priority devices can be stored in solid-state drives to improve access efficiency, while data of low-priority devices can be stored in mechanical hard drives to reduce costs.

[0086] The data encryption and authentication submodule is responsible for ensuring the security of data transmitted and stored in the system and preventing illegal access and data leakage.

[0087] In a possible implementation, the submodule adopts a hybrid encryption mechanism based on symmetric encryption and asymmetric encryption. For example: When the system transmits data, it uses TLS (Transport Layer Security) to encrypt the transmitted content to ensure that the data will not be stolen or tampered with during transmission.

[0088] When storing sensitive data, the system uses AES (Advanced Encryption Standard) to encrypt the data. Specifically, a randomly generated key is attached to each record to prevent its content from being inferred through data patterns.

[0089] As an option, this submodule also provides two-way authentication. Generally, when a device accesses the cloud system, it will complete the identity verification through a digital certificate to ensure that the connected device is a trusted source. At the same time, the cloud will also provide authentication credentials to the device to prevent illegal simulated cloud services from deceiving the device.

[0090] Specifically, the certification process includes the following steps: The device sends a message containing the device ID to the cloud. i Request for certification; The cloud verifies the device’s identity by checking its digital certificate; After verification, the cloud generates a session key K s , used for subsequent communication encryption.

[0091] The privacy protection submodule uses differential privacy technology to protect sensitive data uploaded by the device from leaking the device identity information when it is analyzed and processed.

[0092] Specifically, this submodule adds random noise to the data uploaded by the device to blur sensitive information. For example, for the resource requirement vector of the device The noise processing formula is: in: Resource requirements after adding noise; ∈: random noise, following Laplace distribution Where b is the noise amplitude parameter, which is used to control the strength of privacy protection.

[0093] As an option, this submodule can also perform layered processing on sensitive data. For example: For common data (such as resource allocation results), store them directly; For sensitive data (such as device task type and priority parameters), noise processing is performed before storage, and the processed data is stored separately from the original data.

[0094] To further enhance privacy protection, this submodule supports the federated learning mechanism, in which sensitive data is always stored locally on the device, and the cloud only obtains encrypted model update parameters, thus avoiding direct access to sensitive information on the device.

[0095] In order to improve data management efficiency, this module can combine data compression technology to reduce the data volume through compression algorithms when storing data. For example, for time series data, a compression method based on piecewise linear fitting can be used to represent continuous resource demand records as linear model parameters, thereby significantly reducing storage usage. In addition, the data encryption and authentication submodule can also combine blockchain technology to sign and record data uploaded by the device in an unalterable manner, providing additional protection for data security.

[0096] In this embodiment, the data management and security module provides data-level security for the operation of the system through data storage, encryption authentication and privacy protection functions.

[0097] Although embodiments of the present invention have been shown and described, it will be appreciated by those skilled in the art that various changes, modifications, substitutions and variations may be made to the embodiments without departing from the principles and spirit of the present invention, and that the scope of the present invention is defined by the appended claims and their equivalents.

Claims

1. A cloud system based on the Internet of Things that can quickly access edge devices, characterized in that: include: The edge device access module is used to receive access requests from edge devices and complete device authentication and communication protocol adaptation; The cloud resource monitoring module is used to monitor the resource usage status of the cloud in real time, including the usage of bandwidth, computing power and storage resources; Cloud optimization scheduling module, which is used to dynamically allocate cloud resources based on the resource requirements and priorities of edge devices; Data feedback and adjustment module, used to feed back resource allocation results to edge devices and dynamically adjust the transmission rate of edge devices; The data management and security module is used to store the historical data and allocation plans of edge devices and to provide security protection for data transmission and storage.

2. According to claim 1, a cloud system based on the Internet of Things that can quickly access edge devices is characterized in that: The edge device access module includes: The device authentication submodule is used to complete device authentication based on the unique identification of the edge device to ensure the legitimacy of device access; The communication protocol adapter module is used to support different communication protocols and convert the resource demand information sent by the edge device into a data format that can be processed by the cloud; The demand upload submodule is used to upload the resource demand information and priority weights of edge devices to the cloud optimization scheduling module.

3. The cloud system based on the Internet of Things that can quickly access edge devices according to claim 1, characterized in that: The cloud resource monitoring module includes: The resource status collection submodule is used to collect the bandwidth, computing power and storage resource usage of the cloud system in real time; The device status monitoring submodule is used to monitor the current operating status of edge devices, including resource allocation and transmission status; The historical data recording submodule is used to record the historical resource requirements and allocation plans of edge devices to provide a reference for subsequent optimization.

4. The cloud system based on the Internet of Things that can quickly access edge devices according to claim 1, characterized in that: The cloud optimization scheduling module includes: The state prediction submodule is used to predict future resource demand based on the historical demand data and current resource status of edge devices; The resource allocation submodule is used to dynamically calculate the resource allocation plan according to the priority weight and demand of the edge device; The iterative scheduling submodule is used to dynamically adjust the allocation parameters according to the allocation results to achieve global balance of resources.

5. The cloud system based on the Internet of Things and capable of quickly accessing edge devices according to claim 4, characterized in that: The resource demand prediction step of the state prediction submodule includes: Collect historical resource demand data of edge devices; Generate future resource requirements based on the current state of edge devices combined with random disturbance factors; The prediction results are provided to the resource allocation submodule.

6. The cloud system based on the Internet of Things that can quickly access edge devices according to claim 4, characterized in that: The resource allocation step of the resource allocation submodule includes: Calculate the resource allocation for each device based on the resource requirements, priority weights, and resource status of the edge device; Ensure that all resource allocations meet the total resource constraints of the cloud system; Optimize resource allocation to ensure that the needs of high-priority devices are met first while ensuring fairness for other devices.

7. The cloud system based on the Internet of Things that can quickly access edge devices according to claim 1, characterized in that: The data feedback and adjustment module includes: The resource allocation feedback submodule is used to feed back the resource allocation results generated by the cloud optimization scheduling module to the edge device; The transmission rate adjustment submodule is used to adjust the transmission rate of edge devices to balance the network load when resources are insufficient; The dynamic compensation submodule is used to adjust the allocation scheme to meet the new demand when the demand of edge devices changes.

8. The cloud system based on the Internet of Things capable of quickly accessing edge devices according to claim 7, characterized in that: The dynamic resource compensation step of the dynamic compensation submodule includes: Detect the current resource usage of edge devices; Determine whether device demand has increased or network load has decreased; Recalculate the allocation plan based on the remaining resources in the cloud and adjust the device allocation resources.

9. The cloud system based on the Internet of Things capable of quickly accessing edge devices according to claim 1, characterized in that: The data management and security module includes: The data storage submodule is used to store the resource requirements, allocation schemes and historical data of edge devices; The data encryption submodule is used to encrypt data during data transmission to prevent data leakage; The data authentication submodule is used to verify the legitimacy of edge devices and ensure that the data source is credible.

10. The cloud system based on the Internet of Things capable of quickly accessing edge devices according to claim 9, characterized in that: The privacy protection function of the data management and security module includes the following steps: The sensitive data uploaded by edge devices is processed in a graded manner, and sensitive information is stored separately from general data; Perform differential privacy processing on sensitive data and add random noise to blur the processed results; The processed data is used for resource allocation calculations to ensure the privacy and security of the device.

Images