Research on Multi Kernel Collaborative Scheduling Algorithm for Embedded Operating Systems Based on Dynamic Load Balancing

In embedded operating systems, as the number of kernels increases, the granularity of task allocation needs to become finer and finer. However, overly detailed task allocation may lead to complex dependencies between tasks, increasing the difficulty of scheduling and directly affecting the effectiveness of load balancing. Therefore, a multi kernel collaborative scheduling algorithm for embedded operating systems based on dynamic load balancing is proposed. By obtaining the load weights of embedded operating system multi-core, selecting schedulable nodes of embedded operating system multi-core, and prioritizing task allocation to cores with lighter loads, dynamic load balancing can be achieved. On this basis, the embedded operating system can reorganize multi-core schedulable node tasks, dynamically generate the number of task items based on node performance, consider factors such as task characteristics, system requirements, and multi-core processor characteristics, and ensure that tasks can be efficiently co scheduled on multiple cores, thereby achieving multi-core co scheduling in embedded operating systems. The experimental results show that the proposed method has good load balancing performance in embedded operating systems and can effectively improve the success rate and efficiency of multi-core collaborative scheduling in embedded operating systems.