Search

CN-121984729-A - IPv 6-based information creation edge node task processing method and system

CN121984729ACN 121984729 ACN121984729 ACN 121984729ACN-121984729-A

Abstract

The invention relates to the field of cloud computing, and provides an IPv 6-based task processing method and system for an information creation edge node. The method comprises the steps of conducting layered division on computing tasks, quantifying resource consumption, protecting user privacy data, jointly optimizing task allocation strategies to obtain a task execution scheme of privacy protection, conducting collaborative training on distributed nodes through a federal learning framework based on data characteristics of a plurality of subtasks to enable model parameters between a central node and edge nodes to interact to obtain node training state information, deploying an edge proxy framework supporting an IPv6 protocol according to the node training state information, determining a transmission path of the subtasks through a source routing mechanism, conducting identification processing on data packets through an expansion header to obtain a task execution result, and conducting quantification scoring on node computing capacity through a multi-dimensional evaluation index system to obtain a node capacity evaluation value. The invention ensures the privacy of the user and improves the efficiency of data processing.

Inventors

  • CHEN YANG

Assignees

  • 天翼物联科技有限公司

Dates

Publication Date
20260505
Application Date
20260120

Claims (10)

  1. 1. The IPv 6-based information creation edge node task processing method is characterized by comprising the following steps: s1, carrying out hierarchical division on computing tasks and quantifying resource consumption, protecting user privacy data through a privacy computing technology, and jointly optimizing task allocation strategies to obtain a task execution scheme with privacy protection; S2, based on data characteristics of a plurality of subtasks in the task execution scheme, performing cooperative training on the distributed nodes through a federal learning framework so as to enable model parameters between the central node and the edge nodes to interact, and obtaining a global optimization model and corresponding node training state information; S3, according to the node training state information, deploying an edge proxy architecture supporting an IPv6 protocol, determining a transmission path of a subtask in a task execution scheme through a source routing mechanism, and identifying and processing a data packet by utilizing an extension header to obtain a task execution result; And S4, quantitatively scoring the node computing capacity by adopting a multi-dimensional evaluation index system based on the task execution result, and obtaining a node capacity evaluation value through normalization processing.
  2. 2. The method for processing tasks of an IPv 6-based source edge node according to claim 1, wherein the specific process of step S1 includes: S11, dividing a computing task of a local mobile device into a plurality of subtasks according to a hierarchical structure, marking execution position attributes of the subtasks to obtain a subtask set comprising local execution subtasks and unloading execution subtasks; s12, acquiring CPU power supply voltage, working frequency and equivalent load capacitance when a plurality of subtasks in the subtask set are executed, and carrying out quantitative evaluation on the power calculation power consumption of the plurality of subtasks based on a power consumption calculation formula to obtain a task power consumption quantitative result; S13, identifying a high-power-consumption task requiring privacy protection according to the task power consumption quantification result, constructing a trust chain from a hardware layer to an application layer by utilizing a trusted computing technology, transmitting a trust relationship to a privacy computing framework layer through an integrity measurement function, and storing an integrity measurement value in a hardware security module to obtain trusted execution environment configuration information; s14, under the constraint of the trusted execution environment configuration information, the constraint conditions of the task queue length, the reasoning result return queue length and the channel condition change are synthesized, and the unloading strategy and the privacy protection strategy of the subtask set are jointly optimized, so that the aim of minimizing the total cost of the system and maximizing the privacy protection of the user is achieved, and a task execution scheme with the privacy protection is obtained.
  3. 3. The method for processing tasks of an IPv6 based source edge node according to claim 2, wherein step S12 further comprises: S121, collecting real-time power consumption data of intelligent computing nodes under different task loads when executing the subtask set; S122, calculating the power consumption deviation degree of the CPU power consumption from the reference value according to the calculation force change or the network flow sudden increase and decrease condition when the subtask set is executed; And S123, when the power consumption deviation exceeds a preset threshold, marking the corresponding subtask as a power consumption abnormal task unit, optimally controlling the power consumption by reducing the power supply voltage or adjusting the working frequency, and updating the task power consumption quantification result.
  4. 4. The method for processing tasks of an IPv6 based source edge node according to claim 2, wherein step S13 further comprises: s131, establishing a hardware trust root when the system is started by utilizing a trusted computing instruction embedded in the CPU; s132, upwards transmitting the trust relationship from the CPU layer to the operating system layer, and starting an IMA integrity measurement function by expanding a Linux operating system; S133, expanding a trust chain to a privacy computing framework of an application layer, and ensuring that an execution module, a user-defined operator and user data of a plurality of subtasks in the subtask set are maintained in the trust chain; s134, verifying the integrity of each hierarchical component through a remote authentication technology, and generating trusted execution environment configuration information.
  5. 5. The method for processing tasks of an IPv 6-based source edge node according to claim 1, wherein step S2 further comprises: S21, initializing a global model by a central node according to the data characteristics of the subtask set and the target application requirements, and distributing the global model to each edge node terminal; S22, training a global model by using local data of corresponding subtasks in the subtask set by a plurality of edge node terminals, and continuously optimizing to local model parameters with minimum loss functions through a gradient descent algorithm; S23, returning the updated local model parameters and training loss values to the central node by the plurality of edge nodes, and carrying out weighted aggregation on the local model parameters of the plurality of nodes by the central node according to a criterion of minimizing the overall average loss to obtain aggregated model parameters; And S24, distributing the global model fused with the aggregate model parameters to a plurality of edge nodes again, and repeating iterative optimization until the model converges to obtain a global optimization model and node training state information, wherein the node state information comprises a plurality of edge node training loss values, iteration times and model convergence states.
  6. 6. The method for processing tasks of an IPv6 based source edge node according to claim 1, wherein step S3 further comprises: S31, according to the model convergence state of a plurality of edge nodes in the node training state information, deploying an edge Agent framework supporting an IPv6 source host, wherein the edge Agent framework comprises a plurality of management nodes responsible for discovery, registration, monitoring and task issuing of the edge nodes, and the edge node Agent is used as an independent computing unit to receive and execute tasks; S32, determining optimal task unloading paths of a plurality of subtasks in the subtask set from the source host to the target node by the source host according to training loss values and iteration times of a plurality of edge nodes in the node training state information and combining the link states and load conditions between the gateway and the edge nodes; S33, generating a unique identifier for each subtask data packet to be transmitted in a segmented mode in the subtask set by a source host, and setting a next header field in a data segment header as a first preset value; And S34, after the edge node finishes the calculation of the corresponding subtasks in the subtask set, packaging the calculation result into options in a destination option header for transmission, discarding task data which are participated in calculation, indicating that the final destination node is reached when the next header field value is a second preset value, and collecting the calculation results of all the subtasks to obtain a task execution result.
  7. 7. The method for processing tasks of an IPv6 based source edge node according to claim 6, wherein step S32 further comprises: s321, collecting network performance indexes of a plurality of transmission links; S322, acquiring node state indexes of a plurality of edge nodes in the node training state information; S323, integrating the network performance index, the node state index and the data quantity of a plurality of subtasks in the subtask set, and calculating the integrated cost value of a plurality of candidate paths through a path selection algorithm; s324, selecting a path with the minimum comprehensive cost value as an optimal task unloading path, and designating the path by a source host to transmit the corresponding subtasks in the subtask set.
  8. 8. The method for processing tasks of an IPv6 based source edge node according to claim 1, wherein step S4 further comprises: S41, selecting a plurality of evaluation indexes, and performing multidimensional evaluation on the data quality of the edge nodes completing the task to obtain an evaluation result; S42, configuring an evaluation index weight according to the application scene demand; s43, calculating an initial scoring value of the edge node according to the evaluation index weight and the discretized evaluation result; s44, carrying out normalization processing on the initial scoring value by adopting a sigmoid operator to obtain an edge node capability evaluating value.
  9. 9. The method according to claim 8, wherein the plurality of evaluation indexes in the step S41 include a data integrity index, a data normalization index, a data consistency index, a data accuracy index, a data uniqueness index and a timeliness index.
  10. 10. An IPv6 based credit edge node task processing system for executing an IPv6 based credit edge node task processing method according to any one of claims 1 to 9, comprising: The quantization module is used for carrying out hierarchical division on the computing tasks and quantizing resource consumption, protecting the user privacy data through a privacy computing technology, and jointly optimizing a task allocation strategy to obtain a task execution scheme with privacy protection; the training module is used for carrying out cooperative training on the distributed nodes through the federal learning framework based on the data characteristics of a plurality of subtasks in the task execution scheme so as to enable model parameters between the central node and the edge nodes to be interacted, and a global optimization model and corresponding node training state information are obtained; the processing module is used for deploying an edge proxy architecture supporting an IPv6 protocol according to the node training state information, determining a transmission path of a subtask in a task execution scheme through a source routing mechanism, and identifying and processing a data packet by utilizing an extension header to obtain a task execution result; and the evaluation module is used for quantitatively scoring the node computing capacity by adopting a multi-dimensional evaluation index system based on the task execution result, and obtaining a node capacity evaluation value through normalization processing.

Description

IPv 6-based information creation edge node task processing method and system Technical Field The invention belongs to the technical field of cloud computing, and particularly relates to an IPv 6-based task processing method and system for an edge node of a message creation. Background With the IPv6 protocol becoming the internet core architecture foundation, data security has become a focus of attention in the digital age. In the context of rapid development of information technology, the wide application of cloud computing and internet of things has led to a dramatic increase in the amount of data generated by users, and edge computing has emerged as an emerging computing paradigm. The core idea of edge computing is to move the corresponding functions of the computing platform from the network core side to the network access side, and deploy nodes with certain computing power on the edge side close to the data source to provide nearby services for users, so as to reduce the delay of end users and enhance the user experience. In the existing edge computing technology, a user usually offloads computing tasks to an edge server to complete various computing tasks. However, the prior art has the defects that firstly, the computing power resources of the edge nodes are limited, when a plurality of computing tasks are simultaneously unloaded to an edge server, resource bottleneck problems are easy to occur, so that the task execution efficiency is low, secondly, the existing task unloading scheme lacks an effective protection mechanism for user privacy data, is easy to be threatened by security when sensitive data are processed on edge equipment, and further, the traditional task unloading strategy determines a forwarding path by a router on a path through which a data packet passes, so that flexibility and controllability are lacking, the task allocation cannot be dynamically optimized according to the real-time load state and network condition of the edge nodes, and finally, the prior art lacks a quantitative evaluation mechanism for the computing capacity of the edge nodes, so that the task cannot be reasonably allocated according to the actual processing capacity of the nodes, and the overall performance of the system is difficult to optimize. Therefore, on the premise of ensuring the privacy safety of users, the realization of intelligent scheduling and efficient execution of the edge node tasks becomes a technical problem to be solved in the field. Disclosure of Invention In view of the above shortcomings in the prior art, an object of the present invention is to provide a method and a system for processing tasks of an IPv 6-based source edge node. The invention provides an IPv 6-based task processing method for an edge node of a message creation, which comprises the following steps: s1, carrying out hierarchical division on computing tasks and quantifying resource consumption, protecting user privacy data through a privacy computing technology, and jointly optimizing task allocation strategies to obtain a task execution scheme with privacy protection; S2, based on data characteristics of a plurality of subtasks in the task execution scheme, performing cooperative training on the distributed nodes through a federal learning framework so as to enable model parameters between the central node and the edge nodes to interact, and obtaining a global optimization model and corresponding node training state information; S3, according to the node training state information, deploying an edge proxy architecture supporting an IPv6 protocol, determining a transmission path of a subtask in a task execution scheme through a source routing mechanism, and identifying and processing a data packet by utilizing an extension header to obtain a task execution result; And S4, quantitatively scoring the node computing capacity by adopting a multi-dimensional evaluation index system based on the task execution result, and obtaining a node capacity evaluation value through normalization processing. According to the method for processing the task of the source edge node based on the IPv6, the specific process of the step S1 comprises the following steps: S11, dividing a computing task of a local mobile device into a plurality of subtasks according to a hierarchical structure, marking execution position attributes of the subtasks to obtain a subtask set comprising local execution subtasks and unloading execution subtasks; s12, acquiring CPU power supply voltage, working frequency and equivalent load capacitance when a plurality of subtasks in the subtask set are executed, and carrying out quantitative evaluation on the power calculation power consumption of the plurality of subtasks based on a power consumption calculation formula to obtain a task power consumption quantitative result; S13, identifying a high-power-consumption task requiring privacy protection according to the task power consumption quantification result, constructi