基于隐马尔可夫模型的内存碎片检测机制

Memory fragmentation detection mechanism based on hidden Markov model

当前Vx Works操作系统缺少内存碎片的检测机制。通过增加内存统计信息,基于隐马尔可夫模型的检测程序分析出系统中哪些任务可能是造成内存碎片的根源。软件开发人员根据分析结果对可能造成内存碎片的代码进行优化,且优化前后的分析数据表明内存碎片问题得到了有效的改善,可满足嵌入式设备减少内存碎片的需求。

Current Vx Works operating system lacks the detection mechanism for memory fragmentation. By adding statistical information of memory, detection procedure based on Hidden Markov Model（ HMM） analyzed which tasks of the system maybe the root causes of memory fragmentation. Software developers optimized the code which may causes the memory fragmentation according to the analysis results, and the analysis results before and after the optimization show that memory fragmentation problem has been sloved effectively for embedded devices.