Cloud edge-end collaborative scheduling optimization method and system based on dynamic resource portrait

Abstract

The invention discloses a cloud side-end collaborative scheduling optimization method and system based on a dynamic resource portrait, and the method comprises the steps: constructing the dynamic resource portrait of a terminal device, carrying out the time-space sparse field modeling, and generating sparse data; based on the dynamic resource portrait and the sparse data, constructing layered differential privacy protection, and performing zero-knowledge scheduling verification on the data after differential privacy protection; for tasks passing zero-knowledge scheduling verification, cloud side-end collaborative scheduling operation is executed, and hierarchical differential privacy protection is optimized so as to realize efficient allocation of resources; according to the method, the continuous spatio-temporal data stream is discretized into the three-dimensional grid, so that the communication delay is reduced, the task response efficiency is improved, in addition, through a hierarchical differential privacy protection mechanism, the problem of low efficiency of privacy protection can be balanced, a zero-knowledge scheduling verification protocol is introduced, the privacy leakage risk is reduced, and the method is suitable for different scheduling application scenes.

Description

TECHNICAL FIELD

[0001] The present application relates to the technical field of cloud collaborative resource scheduling, and in particular to a cloud-edge-terminal collaborative scheduling optimization method and system based on dynamic resource profiling. BACKGROUND

[0002] In recent years, with the rapid development of Internet of Things (IoT), edge computing and cloud computing, cloud, edge and terminal collaborative scheduling technology has attracted widespread attention in the fields of intelligent transportation, industrial Internet of Things and smart cities. This technology aims to optimize resource allocation and task scheduling through the collaborative work of terminal devices, edge nodes and cloud servers to meet the needs of large-scale, dynamic and heterogeneous distributed systems. In the field of intelligent transportation systems, connected vehicles collect real-time location, speed and task request data through on-board units, edge nodes (such as roadside units) perform regional-level data processing and task allocation, and cloud servers are responsible for global optimization and policy formulation. Among them, related technologies have developed various scheduling methods, such as cloud scheduling based on centralized optimization, distributed scheduling based on edge computing, and hybrid scheduling methods combining machine learning and heuristic algorithms. In addition, privacy protection technology and zero-knowledge proof technology are introduced to protect sensitive data and meet data security and compliance requirements. The progress of these technologies has significantly improved system performance, promoting real-time task processing and resource utilization efficiency, especially in highly dynamic scenarios.

[0003] However, existing technologies still have significant deficiencies in dynamic resource management and privacy protection. First, traditional scheduling methods rely heavily on static resource models, making it difficult to effectively capture the real-time state changes of terminal devices (such as connected vehicles), such as dynamic changes in power, computing capacity and mobile trajectory, resulting in scheduling decisions that cannot adapt to rapid fluctuations in traffic flow or task demand. Second, existing privacy protection mechanisms usually adopt a unified differential privacy policy, ignoring the spatiotemporal heterogeneity of data distribution, making it difficult to maintain sufficient data utility in high-density areas (such as city centers) while providing strong privacy protection in low-density areas (such as suburbs), resulting in inefficient privacy-utility balance. In addition, traditional scheduling verification methods usually need to expose device resource details, increasing the risk of privacy leakage, especially when sensitive tasks (such as emergency rescue) are involved, lacking an efficient privacy protection verification mechanism. SUMMARY

[0004] This section aims to summarize some aspects of the embodiments of the present application and briefly introduce some preferred embodiments. Some simplifications or omissions may be made in this section and the abstract and title of the specification to avoid obscuring the purpose of this section, abstract and title, and such simplifications or omissions cannot be used to limit the scope of the present application.

[0005] In view of the above existing problems, the present application is proposed. Therefore, the present application provides a cloud edge end collaborative scheduling optimization method based on dynamic resource portrait to solve the problems mentioned in the background art.

[0006] To solve the above technical problems, the present application provides the following technical solutions:

[0007] In a first aspect, the present application provides a cloud edge end collaborative scheduling optimization method based on dynamic resource portrait, comprising:

[0008] Constructing a dynamic resource portrait of a terminal device and performing spatio-temporal sparse field modeling to generate sparse data;

[0009] Based on the dynamic resource portrait and sparse data, constructing hierarchical differential privacy protection, and performing zero-knowledge scheduling verification on the data after differential privacy protection;

[0010] For the tasks that pass the zero-knowledge scheduling verification, performing cloud edge end collaborative scheduling operation, optimizing the hierarchical differential privacy protection to realize efficient allocation of resources.

[0011] As a preferred scheme of the cloud edge end collaborative scheduling optimization method based on dynamic resource portrait of the present application, wherein: the constructing a dynamic resource portrait of a terminal device and performing spatio-temporal sparse field modeling to generate sparse data comprises:

[0012] Collecting spatio-temporal data stream containing time stamp, geographic coordinate and task attribute through a terminal device, and discretizing the spatio-temporal data stream into three-dimensional spatio-temporal grid;

[0013] According to the information entropy of data volume, grid volume, data distribution and total number of terminal devices in the three-dimensional spatio-temporal grid, calculating the privacy strength index of each three-dimensional spatio-temporal grid.

[0014] As a preferred scheme of the cloud edge end collaborative scheduling optimization method based on dynamic resource portrait of the present application, wherein: further comprising:

[0015] Pretreating the collected spatio-temporal data stream;

[0016] Dynamically adjusting the size of geographic grid in three-dimensional spatio-temporal grid based on the data density of spatio-temporal data stream;

[0017] Performing adaptive wavelet transform processing on the spatio-temporal grid to generate sparse data.

[0018] As a preferred scheme of the cloud edge end collaborative scheduling optimization method based on dynamic resource portrait of the present application, wherein: based on the dynamic resource portrait and sparse data, constructing hierarchical differential privacy protection comprises:

[0019] At the terminal device layer, the original space-time data stream is protected by clipping and adding noise;

[0020] At the edge node layer, Gaussian noise is added to the aggregated regional features;

[0021] At the cloud layer, the cluster resource scheduling system is disturbed by a probability selection mechanism.

[0022] As a preferred scheme of the cloud edge end collaborative scheduling optimization method based on dynamic resource portrait according to the application, wherein:

[0023] The original space-time data stream is limited in a preset range by clipping;

[0024] The probability selection mechanism determines the disturbance strength based on the reconstruction error and execution efficiency of the original space-time data stream.

[0025] As a preferred scheme of the cloud edge end collaborative scheduling optimization method based on dynamic resource portrait according to the application, wherein:

[0026] The scheduling constraints are converted into an arithmetic circuit, and the constraints include space-time constraints, resource constraints and task priority constraints;

[0027] A zero-knowledge proof is generated at the edge node, and the proof is generated based on the hash value of the arithmetic circuit and the dynamic resource portrait and the private input of the three-dimensional space-time grid coordinates;

[0028] The validity of the zero-knowledge proof is verified by a smart contract at the cloud end, and the tasks that pass the verification are added to the scheduling queue.

[0029] As a preferred scheme of the cloud edge end collaborative scheduling optimization method based on dynamic resource portrait according to the application, wherein:

[0030] At the terminal device layer, the task feasibility is evaluated by local pre-screening, and the evaluation is based on the execution time of the task;

[0031] At the edge node layer, the task allocation within the region is optimized by a sliding window to minimize the weighted sum of the load balancing index and the resource allocation change;

[0032] At the cloud layer, the task allocation is optimized by a genetic algorithm.

[0033] In a second aspect, the application provides a cloud edge end collaborative scheduling optimization system based on dynamic resource portrait, which comprises:

[0034] The dynamic resource modeling and sparse processing module is configured to construct the dynamic resource portrait of the terminal device and perform space-time sparse field modeling to generate sparse data;

[0035] The hierarchical privacy protection and zero-knowledge verification module is configured to construct hierarchical differential privacy protection based on the dynamic resource portrait and the sparsification data, and perform zero-knowledge scheduling verification on the data after differential privacy protection.

[0036] The cloud-edge-end collaborative scheduling and optimization module is configured to perform cloud-edge-end collaborative scheduling operation on the task that passes the zero-knowledge scheduling verification, and optimize the hierarchical differential privacy protection to realize efficient allocation of resources.

[0037] In a third aspect, the present application provides a computer device comprising a memory and a processor, wherein the memory stores a computer program, and wherein the processor implements any step of the above method when executing the computer program.

[0038] In a fourth aspect, the present application provides a computer readable storage medium having a computer program stored thereon, wherein the computer program, when executed by a processor, implements any step of the above method.

[0039] Compared with the prior art, the present application has the following advantages:

[0040] 1. By constructing the dynamic resource portrait of the terminal device, the present application can capture the dynamic changes of the device state in real time, and combine the spatiotemporal sparse field modeling to discretize the continuous spatiotemporal data stream into a three-dimensional grid, thereby reducing communication delay and improving the efficiency of task response.

[0041] 2. The present application adds strategies at the terminal, edge and cloud end through the hierarchical differential privacy protection mechanism, retains data details in high-density areas to support high-precision tasks, and enhances privacy protection in low-density areas, thereby balancing the low-efficiency problem of privacy protection.

[0042] 3. By introducing the zero-knowledge scheduling verification protocol, the scheduling constraints are converted into arithmetic circuits, zero-knowledge proofs are generated at the edge node, and the cloud end is verified through the smart contract, thereby ensuring that the task allocation meets the spatiotemporal constraints and resource constraints without exposing the sensitive details of the dynamic resource portrait, thereby reducing the risk of privacy leakage. BRIEF DESCRIPTION OF DRAWINGS

[0043] In order to more clearly illustrate the technical solutions of the embodiments of the present application, the following will briefly introduce the drawings needed in the embodiment description. Obviously, the drawings in the following description are only some embodiments of the present application, and for those skilled in the art, other drawings can also be obtained without creative labor. Among them:

[0044] Figure 1 The overall flowchart of the cloud-edge-end collaborative scheduling optimization method based on dynamic resource portrait according to an embodiment of the present application is shown. DETAILED DESCRIPTION

[0045] In order to make the above objectives, features and advantages of the present application more clear and easily understood, the specific embodiments of the present application will be described in detail below with the accompanying drawings. Obviously, the described embodiments are only a part of the embodiments of the present application, rather than all the embodiments. Based on the embodiments in the present application, all the other embodiments obtained by those skilled in the art without creative work should belong to the protection scope of the present application.

[0046] In the following description, numerous specific details are set forth in order to provide a thorough understanding of the present application. However, it will be apparent to one skilled in the art that the present application can be practiced without the specific details presented herein. In other instances, well-known methods have not been described in detail in order to avoid obscuring aspects of the present application.

[0047] Secondly, the "one embodiment" or "embodiment" referred to herein means that the specific features, structures or characteristics can be included in at least one implementation of the present application. The "in one embodiment" appearing in different places in the specification does not mean the same embodiment, nor is it an independent or alternative embodiment mutually exclusive with other embodiments.

[0048] The present application is described in detail in conjunction with the schematic diagram. In the detailed description of the embodiments of the present application, the sectional view of the device structure is locally enlarged without the general proportion for the convenience of description, and the schematic diagram is only an example, which should not limit the protection scope of the present application herein. In addition, the three-dimensional spatial dimensions of length, width and depth should be included in the actual manufacture.

[0049] Meanwhile, in the description of the present application, it should be noted that the terms "upper, lower, inner and outer" and the like indicate the orientation or positional relationship based on the orientation or positional relationship shown in the drawings, and are only for the convenience of describing the present application and simplifying the description, and do not indicate or imply that the indicated device or element must have a particular orientation, be constructed and operated in a particular orientation, and therefore cannot be understood as a limitation on the present application. In addition, the terms "first, second or third" are only for the purpose of description, and cannot be understood as indicating or implying relative importance.

[0050] In the present application, unless otherwise explicitly specified and limited, the terms "mounting, connecting, connection" should be understood broadly, for example: it can be fixed connection, detachable connection or integral connection; it can also be mechanical connection, electrical connection or direct connection, it can also be indirectly connected through intermediate medium, or it can be the communication inside two elements. For those skilled in the art, the specific meaning of the above terms in the present application can be understood according to the specific circumstances.

[0051] Example 1

[0052] Referring to Figure 1 For the first embodiment of the present application, the embodiment provides a cloud edge terminal collaborative scheduling optimization method based on dynamic resource portrait, comprising:

[0053] S1, constructing a dynamic resource portrait of a terminal device and performing space-time sparse field modeling to generate sparse data;

[0054] It should be explained that the dynamic resource portrait is a multi-dimensional feature extraction of the real-time state of the terminal device, edge node and cloud resource, forming a structured feature vector for supporting scheduling decision;

[0055] Specifically, the resource portrait includes computing power (CPU / GPU performance), storage capacity, communication bandwidth, energy consumption state and space-time location, etc.

[0056] Further, for the space-time distribution of the modeled resource portrait, the present application adopts a space-time sparse field method to discretize the continuous space-time data stream into a three-dimensional grid (time, longitude, latitude), and assumes that the space-time data stream collected by the terminal device is Where t is the time stamp, (x, y) is the geographic coordinate, i represents the number of terminal devices, n represents the total number of terminal devices, s i is the task attribute;

[0057] Specifically, the collected space-time data stream is preprocessed, which adopts Min-Max normalization to unify the dimension of the space-time data stream and constrain its range to [0, 1]. Since this preprocessing operation is a conventional technical means for data processing, it will not be described here.

[0058] In addition, in order to make the present application clear and clear, the intelligent transportation system is used as an example in this embodiment, but it is not limited to the intelligent transportation system. It is assumed that a city has thousands of networked vehicles, hundreds of roadside units and a central cloud server. The vehicle data is collected through the vehicle unit, and the construction process of the dynamic resource portrait is as follows: collecting vehicle data uploaded to the roadside unit every second, the data including time, location and task type (such as emergency rescue or regular navigation), dividing the city area into space grid and time slice, each space grid is a three-dimensional grid, then the resource portrait of each vehicle includes static features (such as task type) and dynamic features (such as time, location);

[0059] Specifically, the continuous space-time data stream is discretized into a three-dimensional grid formula as follows:

[0060]

[0061] Where G k,l,mrepresents a three-dimensional grid for organizing and storing the set of data points generated by terminal devices in time and geographical area, each grid is a data container, which contains multiple spatio-temporal data streams (Also referred to as data points in the grid) For example, in the intelligent transportation system, G k,l,m represents all vehicle data (such as location, speed, task request, etc.) in a certain 5-minute time period (such as 18:00:00 to 18:05:00) and a certain 500m x 500m geographical area (such as the city center intersection), and s i (t, x, y) is embodied as vehicle data; k represents a discrete index in the time dimension, which is used to define the time interval [kΔt, (k+1)Δt), through which the data points in the same time period can be aggregated to form the time dimension of the spatio-temporal grid, and Δt is the preset time slicing length (such as 5 minutes), for example, in the intelligent transportation system, if Δt = 300s (5 minutes), then k = 1 represents the first time slice (0 to 300 seconds), and k = 2 represents the second time slice (300 to 600 seconds); (x, y) ∈ C l,m represents that the geographical coordinates of the data points belong to the geographical grid C l,m defined area, which is formed by dividing the geographical area into fixed-size grids, such as the above (a certain 500m x 500m geographical area), the geographical grid defines the spatial range of the data points, facilitating the analysis of vehicle distribution and task demand in the geographical area; U represents the union operator, which combines the elements of multiple sets into a set, containing all non-repeating elements, in the formula, first all data points in the geographical grid C l,m are combined into a set, then t∈[kΔt,(k+1)Δt) all data point sets in the time interval [kΔt, (k+1)Δt) are further combined, and finally, G k,l,m is formed, which contains all data points of terminal devices in the time slice and geographical grid;

[0062] It should be noted that the continuous time is discretized by the time slicing index k, which facilitates the organization of data by time period to analyze the dynamic changes of data flow;

[0063] Further, according to the amount of data in the three-dimensional spatio-temporal grid, the volume of the grid, the information entropy of data distribution, and the total number of terminal devices, the privacy strength index Φ(G) of each three-dimensional spatio-temporal grid is calculated:

[0064]

[0065] Where ||G||1 is the amount of data in the three-dimensional spatio-temporal grid, V is the volume of the grid, and H(G) is the information entropy of data distribution;

[0066] It needs to be explained that, reflects the concentration degree of data points in the grid, and high-density areas (such as city centers) usually have higher privacy sensitivity, because there are more data points, so individual information is easier to infer; and reflects the diversity of data attributes measured by information entropy, the higher the entropy, the more uniform the data distribution, the more difficult it is to infer individual information, and the higher the privacy strength; therefore, the two play a balancing role in privacy strength;

[0067] Further, based on the data density of the space-time data stream, the size of the geographic grid in the three-dimensional space-time grid is dynamically adjusted;

[0068] Specifically, the data density is obtained by the ratio of the total number of data points in the geographic grid to the area of the geographic grid within a certain time slice, and the data density changes dynamically with the time slice, and is recalculated every minute to reflect the real-time changes of traffic flow;

[0069] It should be noted that the time slice is usually not directly used for data density calculation, but the length of the time slice can be additionally considered in the calculation of the privacy strength index to obtain the data privacy strength at a specific time;

[0070] Specifically, according to the dynamic change of data density, the size of the geographic grid is dynamically adjusted by the quadtree algorithm, that is, in high-density areas (such as city centers), the current geographic grid size is refined (reduced) to capture finer vehicle distribution, and in low-density areas (such as suburbs), the current geographic grid size is relaxed (expanded) to reduce the computational overhead of modeling resource profiling;

[0071] Further, the space-time grid is subjected to adaptive wavelet transform processing to generate sparse data:

[0072]

[0073] Among them, denotes the sparse data after adaptive wavelet transform processing, which is a compressed representation of G k,l,m , W(·) denotes the wavelet transform operation, which converts G k,l,m from the time domain to the wavelet domain, decomposing it into low-frequency (such as the main trend of vehicle distribution) and high-frequency components (such as the small fluctuations of vehicle trajectories), W -1 (·) denotes the inverse wavelet transform operation, which converts the coefficients after threshold processing from the wavelet domain back to the original domain to generate sparse data; Thresh(·,λ·Φ(G)) denotes the threshold processing function, which selects the coefficients after wavelet transform, retains the coefficients whose absolute value is greater than the threshold λ·Φ(G), and discards the coefficients less than the threshold to achieve data compression, λ denotes the threshold adjustment factor, which is inversely proportional to the privacy strength index Φ(G).

[0074] It should be explained that through adaptive wavelet transform processing, the amount of compressed data is reduced, and sparse data is generated, which can reduce the problem of edge node (such as roadside unit) and communication bandwidth occupation;

[0075] S2, based on dynamic resource image and sparse data, construct hierarchical differential privacy protection, and perform zero-knowledge scheduling verification on the data after differential privacy protection;

[0076] It should be explained that although the aforementioned only mentions the balance of privacy strength, there may still be privacy leakage problems, so on this basis, a privacy protection mechanism needs to be established;

[0077] Further, a three-layer differential privacy protection mechanism is constructed, at the terminal device layer, the original space-time data stream is protected by cropping and adding noise; at the edge node layer, Gaussian noise is added to the aggregated regional features; at the cloud layer, the cluster resource scheduling system is disturbed through a probability selection mechanism;

[0078] Specifically, taking vehicle location and task request (such as emergency rescue) as an example, Laplace noise is added to the vehicle location and task type data, for example, a higher noise is assigned to a regular navigation task, and a lower noise is assigned to an emergency rescue task, in addition, considering the maximum possible movement distance of vehicle location data and the maximum request change of task type data, the data cropping range is set to ensure the rationality of the data, for example, the vehicle location that exceeds the range is cropped to the city boundary, and then the vehicle location and task type data after adding noise are uploaded to the roadside unit through an encrypted channel (such as TLS), wherein the roadside unit aggregates vehicle data to generate regional level features (such as average speed, flow density), and adds Gaussian noise to the features, the noise level is adjusted according to Φ(G), for example, a small noise is assigned to the city center area to retain the analysis accuracy, and a large noise is assigned to the suburbs; the cloud assigns the optimal path to the vehicles in multiple regions through a probability selection mechanism, the probability selection mechanism uses an exponential mechanism, and the probability of being selected is high for high utility parameters (such as the weight of optimizing path planning), and the disturbance is small; the probability is low for low utility parameters, and the disturbance is large; the utility is calculated by reconstructing the error (i.e. comparing the mean square error of the difference between the disturbed data and the original data) and G k,l,m ;

[0079] It should be explained that the disturbance operation is mainly performed by randomizing the selection parameters to introduce uncertainty and prevent attackers from inferring the original data from the scheduling results of the cluster resource scheduling system;

[0080] It needs to be explained that the cluster resource scheduling system as a distributed computing framework on the cloud layer is mainly responsible for managing and optimizing global resource allocation and task scheduling. It coordinates multiple computing nodes through distributed clusters (such as Kubernetes, Apache Mesos) to handle cross-regional task allocation and resource management; in the intelligent transportation system, it mainly optimizes cross-regional tasks (such as path planning, emergency rescue scheduling) according to the sparse data uploaded by terminal devices (such as connected vehicles) and edge nodes (such as roadside units);

[0081] Further, the scheduling task in the differential privacy protection mechanism is verified by zero-knowledge scheduling;

[0082] Further, the scheduling constraints are converted into arithmetic circuits, including time and space constraints, resource constraints, and task priority constraints;

[0083] Specifically, the scheduling constraints refer to the conditions that must be met by task allocation, including time and space constraints: time and space limitations for task execution, such as vehicles must arrive at the target location within 5 minutes and the distance must not exceed 5 kilometers; resource constraints: device resources (such as computing power, power) meet task requirements, such as path planning tasks require 1TFLOPs computing power; task priority constraints: high-priority tasks (such as emergency rescue) are prioritized over low-priority tasks (such as regular navigation);

[0084] It needs to be explained that the arithmetic circuit is a mathematical representation form, which is used to convert the above constraints into a combination of polynomial operations composed of addition and multiplication gates, suitable for zero-knowledge proof protocols (such as zk-SNARK) processing;

[0085] Further, a zero-knowledge proof is generated at the edge node, proving that the private input is generated based on the hash value of the arithmetic circuit and the dynamic resource profile and the three-dimensional space-time grid coordinates;

[0086] Specifically, the hash value of the dynamic resource profile is calculated using encryption hash functions such as SHA-256, and this hash value is used as public input to ensure privacy (original data is not exposed) and verifiability (hash value is unique);

[0087] Specifically, the data point coordinates (t, x, y) are used as private information to represent the vehicle's position within the time slice and geographic grid;

[0088] It should be noted that private input can not directly expose data information, but is only used for zero-knowledge proof generation;

[0089] Specifically, the zero-knowledge proof generation uses the zk-SNARK protocol (such as libsnark), is based on an arithmetic circuit, generates public parameters (i.e., proof keys and verification keys) in the cloud layer, inputs each constraint parameter in the scheduling constraint and the hash value of the corresponding constraint parameter through the edge node as public input, and simultaneously inputs the private information as private input; according to the public input and the private input, the zero-knowledge proof is generated for about 100-500 milliseconds (the number of milliseconds matches the processor of the edge node);

[0090] Further, the validity of the zero-knowledge proof is verified through the smart contract in the cloud, and the task that passes the verification is added to the scheduling queue;

[0091] Specifically, the smart contract loads the verification key (obtained from the public parameters generated in the cloud layer in the foregoing zk-SNARK protocol), inputs the parameters in the zero-knowledge proof and the public input, and verifies whether the zero-knowledge proof satisfies the arithmetic circuit (this process does not need to access the private input); the task that passes the proof is assigned a unique ID, added to the scheduling queue, and allocated with cloud resources (such as a GPU computing path) and issued to the edge node and the actual vehicle data;

[0092] S3, for the task that passes the zero-knowledge scheduling verification, a cloud-edge-end collaborative scheduling operation is performed, and a hierarchical differential privacy protection is optimized to realize efficient allocation of resources;

[0093] Further, in the terminal device layer, the feasibility of the task is evaluated through local pre-screening, and the time based on the task execution is evaluated;

[0094] Specifically, the terminal device (such as a connected vehicle) performs preliminary evaluation locally, judges whether the task (such as path planning) is feasible, reduces invalid task uploading, and the standard of preliminary evaluation is determined according to the task execution time (the time required to complete the task does not exceed the maximum allowed time);

[0095] It should be noted that the purpose of pre-screening is to ensure that the task that meets the time is uploaded to the edge node, thereby reducing the communication overhead;

[0096] Further, in the edge node layer, the task allocation in the region is optimized through a sliding window;

[0097] Specifically, the length and sliding interval of the sliding window are set, the task set and the resource set in the region are taken as data representation in the sliding window, the uniformity of task allocation is measured by variance, the load balancing index is determined, the difference between the task allocation of the current window and the task allocation of the previous window is measured by L2 norm, and the resource allocation change is determined;

[0098] Further, the load balancing index (uniformity of task allocation) and resource allocation changes (stability of scheduling adjustment) are weighted and summed, wherein the load balancing index is minimized, a greedy algorithm or linear programming is used to solve within the window, and the solution is distributed to the vehicles to ensure that each vehicle is balanced in tasks and frequent adjustments are avoided.

[0099] Further, at the cloud layer, the task allocation is optimized by a genetic algorithm.

[0100] Specifically, a target is established using a genetic algorithm to coordinate cross-regional task allocation to maximize global resource utilization and achieve an optimal allocation and scheduling scheme within a city.

[0101] Specifically, the genetic algorithm generates an initial population (allocation scheme), each allocation scheme is a mapping of tasks to resources, and then generates fragments (individuals) in the allocation scheme through its own selection, crossover and mutation operations. When the maximum number of iterations of the genetic algorithm is reached, the optimal allocation and scheduling scheme is output.

[0102] Further, the embodiment also provides a cloud-edge-end collaborative scheduling optimization system based on a dynamic resource portrait, comprising:

[0103] The dynamic resource modeling and sparsification processing module is configured to construct a dynamic resource portrait of a terminal device and perform spatiotemporal sparse field modeling to generate sparsified data.

[0104] The hierarchical privacy protection and zero-knowledge verification module is configured to construct hierarchical differential privacy protection based on the dynamic resource portrait and the sparsified data, and perform zero-knowledge scheduling verification on the data after differential privacy protection.

[0105] The cloud-edge-end collaborative scheduling and optimization module is configured to perform cloud-edge-end collaborative scheduling operations for tasks that pass the zero-knowledge scheduling verification, and optimize the hierarchical differential privacy protection to achieve efficient allocation of resources.

[0106] The embodiment also provides a computer device suitable for the cloud-edge-end collaborative scheduling optimization method based on a dynamic resource portrait, comprising:

[0107] The memory is used to store computer executable instructions, and the processor is used to execute the computer executable instructions to implement the cloud-edge-end collaborative scheduling optimization method based on a dynamic resource portrait as proposed in the above embodiment.

[0108] The computer device can be a terminal, and the computer device includes a processor, a memory, a communication interface, a display screen and an input device connected by a system bus. The processor of the computer device is configured to provide computing and control capabilities. The memory of the computer device includes a non-volatile storage medium and an internal memory. The non-volatile storage medium stores an operating system and a computer program. The internal memory provides an environment for running the operating system and the computer program in the non-volatile storage medium. The communication interface of the computer device is configured to perform wired or wireless communication with an external terminal. The wireless communication can be achieved by WIFI, an operator network, NFC (Near Field Communication) or other technologies. The display screen of the computer device can be a liquid crystal display screen or an electronic ink display screen. The input device of the computer device can be a touch layer overlaid on the display screen, or a key, trackball or touchpad arranged on the shell of the computer device, or an external keyboard, touchpad or mouse.

[0109] The embodiment also provides a storage medium having a computer program stored thereon, and the computer program is executed by a processor to implement the cloud-edge-terminal collaborative scheduling optimization method based on a dynamic resource portrait as described in the above embodiment.

[0110] The storage medium provided in the embodiment belongs to the same inventive concept as the data storage method provided in the above embodiment. The technical details not described in the embodiment can be referred to the above embodiment, and the embodiment has the same beneficial effects as the above embodiment.

[0111] Those skilled in the art should understand that the embodiments of the present application can be provided as a method, a system or a computer program product. Therefore, the present application can be in the form of an entirely hardware embodiment, an entirely software embodiment or an embodiment combining software and hardware. In addition, the present application can be in the form of a computer program product implemented on one or more computer-usable storage media (including but not limited to disk storage, CD-ROM, optical storage, etc.) containing computer-usable program code. The solutions in the embodiments of the present application can be implemented in various computer languages, such as object-oriented programming language Java and interpreted scripting language JavaScript.

[0112] The present application is described with reference to flowcharts and / or block diagrams according to the methods, devices (systems) and computer program products of the embodiments of the present application. It should be understood that each flow and / or block in the flowcharts and / or block diagrams, and the combination of the flows and / or blocks in the flowcharts and / or block diagrams can be implemented by computer program instructions. These computer program instructions can be provided to a processor of a general-purpose computer, a special-purpose computer, an embedded processor or other programmable data processing device to produce a machine, so that the instructions executed by the processor of the computer or other programmable data processing device produce a device for implementing the functions described in the flowcharts and / or block diagrams.Figure 1 one or more processes and / or blocks Figure 1 one or more blocks that can, working together, implement the

[0113] These computer program instructions can also be stored in a computer readable memory that can direct a computer or other programmable data processing apparatus to function in a Figure 1 one or more processes and / or blocks Figure 1 one or more blocks that can, working together, implement the

[0114] These computer program instructions can also be loaded onto a computer or other programmable data processing apparatus to cause a series of operational steps to be performed on the computer or other programmable apparatus to produce a computer implemented process such that the instructions which execute on the computer or other programmable apparatus provide steps for implementing the Figure 1 one or more processes and / or blocks Figure 1 one or more blocks that can, working together, implement the

[0115] Although preferred embodiments of the application have been described herein, changes and modifications can be suggested to one skilled in the art and it is intended that the application encompass such changes and modifications as fall within the scope of the appended claims. The patent claims recited herein are intended to encompass at least the following embodiments.

[0116] Obviously, numerous modifications and variations of the present application are possible in light of the above teachings. It is therefore to be understood that within the scope of the appended claims and their equivalents, the application can be practiced otherwise than as specifically described herein.

Claims

1. A cloud-edge-device collaborative scheduling optimization method based on dynamic resource profiling, characterized in that: include: Build dynamic resource profiles of terminal devices and perform spatiotemporal sparse field modeling to generate sparse data; Based on the dynamic resource profile and sparse data, a hierarchical differential privacy protection is constructed, and the differential privacy protected data is subjected to zero-knowledge scheduling verification; For tasks that have passed the zero-knowledge scheduling verification, cloud-edge collaborative scheduling operations are performed, and the layered differential privacy protection is optimized to achieve efficient resource allocation.

2. The cloud-edge-device collaborative scheduling optimization method based on dynamic resource profiling according to claim 1 is characterized in that: The process of constructing a dynamic resource profile of a terminal device and performing spatiotemporal sparse field modeling to generate sparse data includes: Collecting a spatiotemporal data stream containing a timestamp, geographic coordinates, and task attributes through a terminal device, and discretizing the spatiotemporal data stream into a three-dimensional spatiotemporal grid; The privacy strength index of each three-dimensional space-time grid is calculated based on the amount of data in the three-dimensional space-time grid, the grid volume, the information entropy of the data distribution, and the total number of terminal devices.

3. The cloud-edge-device collaborative scheduling optimization method based on dynamic resource profiling according to claim 2 is characterized in that: Also includes: Preprocessing the collected spatiotemporal data stream; Dynamically adjust the geographic grid size in a three-dimensional space-time grid based on the data density of the space-time data stream; Adaptive wavelet transform processing is performed on the spatiotemporal grid to generate sparse data.

4. The cloud-edge-device collaborative scheduling optimization method based on dynamic resource profiling according to claim 1 or 3 is characterized in that: Based on the dynamic resource profile and sparse data, a hierarchical differential privacy protection is constructed, including: At the terminal device layer, the original spatiotemporal data stream is protected by cropping and adding noise; At the edge node layer, Gaussian noise is added to the aggregated regional features; At the cloud layer, the cluster resource scheduling system is disturbed through a probabilistic selection mechanism.

5. The cloud-edge-device collaborative scheduling optimization method based on dynamic resource profiling according to claim 4 is characterized in that: Also includes: The original spatiotemporal data stream is restricted to a preset range by clipping; The probabilistic selection mechanism determines the perturbation intensity based on the reconstruction error and execution efficiency of the original spatiotemporal data stream.

6. The cloud-edge-device collaborative scheduling optimization method based on dynamic resource profiling according to claim 1 is characterized in that: The zero-knowledge scheduling verification includes: converting scheduling constraints into arithmetic circuits, wherein the constraints include time and space constraints, resource constraints, and task priority constraints; Generate a zero-knowledge proof at the edge node based on a hash of the arithmetic circuit and the dynamic resource profile, and private input of the three-dimensional space-time grid coordinates; The validity of the zero-knowledge proof is verified through a smart contract in the cloud, and the tasks that pass the verification are added to the scheduling queue.

7. The cloud-edge-device collaborative scheduling optimization method based on dynamic resource profiling according to claim 1 is characterized in that: Perform cloud-edge-device coordinated scheduling operations, including: At the terminal device level, task feasibility is assessed through local pre-screening, where the assessment is based on the time it takes to execute the task; At the edge node layer, a sliding window is used to optimize task allocation within the region, minimizing the weighted sum of load balancing metrics and resource allocation changes. At the cloud layer, task allocation is optimized using genetic algorithms.

8. A cloud-edge-device collaborative scheduling optimization system based on dynamic resource profiling, based on the cloud-edge-device collaborative scheduling optimization method based on dynamic resource profiling according to any one of claims 1 to 7, characterized in that: include: The dynamic resource modeling and sparse processing module is configured to construct a dynamic resource profile of the terminal device and perform spatiotemporal sparse field modeling to generate sparse data; A layered privacy protection and zero-knowledge verification module is configured to construct a layered differential privacy protection based on the dynamic resource profile and the sparse data, and perform zero-knowledge scheduling verification on the differentially privacy protected data; The cloud-edge collaborative scheduling and optimization module is configured to perform cloud-edge collaborative scheduling operations for tasks that have passed the zero-knowledge scheduling verification, optimize the hierarchical differential privacy protection, and achieve efficient resource allocation.

9. A computer device comprising a memory and a processor, wherein the memory stores a computer program, wherein: When the processor executes the computer program, the steps of the method according to any one of claims 1 to 7 are implemented.

10. A computer-readable storage medium having a computer program stored thereon, characterized in that: When the computer program is executed by a processor, the steps of the method according to any one of claims 1 to 7 are implemented.

Images