Research On Real-time Task Scheduling Of Micro-kernel Operating System Of Heterogeneous Multi-core Processor

Heterogeneous multi-core processor is the main trend of multi-core development.Its real-time scheduling problem has become a hot topic in the field of computer science.Based on the scheduling optimization model of heterogeneous multi-core processors,this paper improves the scheduling algorithm for real-time task scheduling.The essence is to use the performance advantage of heterogeneous multi-core processors to solve the problem of task allocation between heterogeneous processors,and to improve the system performance of the micro-kernel operating system.In this paper,the main tasks of real-time task scheduling for heterogeneous multi-core processors are:(a)Firstly In this paper,the modeling design of task model,multi-core processor model,scheduling model and constraint condition model is carried out based on the heterogeneous multi-core system model,which provides a model target environment for the optimization of the algorithm.In order to solve the problem that heterogeneous multi-core processors are difficult to schedule a real-time task with complex dependencies,a parallel processing task tree model(PTT)is designed on the basis of directed acyclic task graph model(DAG).This model defines the priority level by analyzing the dependency between tasks,and determines the requirement level of real-time response of tasks,and then schedules these tasks to corresponding kernels to execute.PTT model mainly solves the problem of an excess expenditure on communication across heterogeneous multi-core tasks,and transforms the task allocation of complex dependency into a task tree model which is convenient to execute.(b)Ant colony algorithm(ACO),which has natural parallel advantages in intelligent algorithm for heterogeneous multi-core system,is used to solve the NP problem in real-time scheduling.The steps of the algorithm design include the selecting operation between tasks,the mapping relationship between tasks and processors,the updating operation of the amount of information,the consideration of cross mutation,the design optimization of the formula and so on.In order to improve the searching efficiency of ant colony algorithm and give full play to the advantages of heterogeneous multi-core processors,it is necessary to optimize the ant colony algorithm that is suitable for heterogeneous multi-core environment.The ant colony algorithm in this paper uses chaos mapping to carry out chaotic disturbance,thus making the basic ant colony algorithm jump out of the local optimal,and designs an improved chaotic ant swarm algorithm(CAS).It can give full play to the heterogeneous and parallel capabilities of heterogeneous multi-core processors.Finally,this paper compares and analyzes the DAG scheduling model and PTT scheduling model under multithreading.The accuracy of the improved scheduling algorithm is verified,and the average running time and average value are compared and analyzed.And compared with other algorithms,the amount of tasks needed for scheduling processing is compared and analyzed.The experimental results show that the PTT model and the improved CAS algorithm are suitable for real-time task scheduling in heterogeneous multi-core of micro-kernel operating system environment,and better scheduling results can be achieved.