Design and Implementation of Dynamic Memory Allocation Algorithm in Embedded Real-Time System

Abstract

With the development of Internet of Things technology, embedded real-time operating system has been more and more widely used. The embedded real-time operating system has higher requirements on the real-time, fragmentation rate and reliability of dynamic memory allocation. Therefore, dynamic memory allocation has become an important research content of embedded real-time operating system. Aiming at the shortage of μC/OS-II memory management mechanism, an improved memory management algorithm is proposed. By predicting transient objects, allocating them on one side of the heap memory, and then allocating the remaining objects on the other side of the heap memory, the algorithm uses enhanced multilevel separation mechanisms and look-up tables and hierarchical bitmaps to make efficient use of memory occupy. The comparison experiment of μC/OS-II platform shows that the improved dynamic memory allocation algorithm can better improve the speed and utilization of memory allocation. The dynamic memory algorithm has better real-time performance and can effectively improve the memory management of embedded real-time operating system performance.