Skip to Main Content
One of the main goals of embedded real-time streaming file system is to reduce the requirement of disk channel and enhance the throughput of file system, as well as to guarantee the real time performance under limited memory environment. Based on the analysis of the real-time streaming file system model, we brought forward the square-root buffer allocation algorithm, which is a statistic allocation algorithm and the rates of IO data streams should be predicted. By reducing the requirement of disk channel utility, the algorithm can improve the throughput of file system. Compared with the proportion allocation algorithm we conclude the algorithm not only guarantee the real time performance of task set under fewer memory condition, but could obviously reduce about 35% the disk channel requirement. So the algorithm is suitable for the embedded application such as the real-time data acquisition systems.