Analysis And Simulation Research Of Linux2.6 Process Scheduling Algorithm

With the rapid development of embedded systems, Linux operating system has been widely used in various fields, such as portable electronic devices, biotechnology, aerospace science and technology and so on. Obviously different applications have different requirements for real-time performance, fairness and throughput of Linux. And the process scheduling algorithm plays a vital role in Linux. Because of the openness of Linux operating system, Linux scheduling algorithm has got a great enrichment and development. Among them, CFS scheduling algorithm is came with the latest Linux 2.6.23.CFS scheduling algorithm has a good application performance on the desktop system, but the performance research of CFS in different applications is still in its beginning stages. So it is urgent to design a simulation method to analyze the scheduling performance of CFS and provide some valuable theoretical guidance for the development of embedded Linux with this method. For this reason, this paper has designed a Linux2.6 process scheduler simulation system and done research on the performance of CFS in different application fields by a series of simulation experiments. In this paper, firstly a theoretical research and analysis on Linux2.6.23 process scheduler, especially the CFS was made. Also, the key algorithms of red-black tree in CFS were researched in this paper. Then the requirement analysis of the Linux2.6 process scheduler simulation system was introduced and a formal model of Linux process scheduler simulation system was set up. And a Linux2.6 process scheduler simulation system was designed and implemented based on this model. A large number of simulation results shown that this simulation system is reliable, stable and easy-to-use. Finally massive simulation experiments on Linux2.6 process scheduler and the latest CFS based on different processors and scheduling parameters were carried out deeply with this simulation system. And several optimization suggestions about CFS in different applications are proposed according to the simulation experiments results and performance ratio. The simulation experiments results showed that the performance of CFS improved with the increase of processor and with the increase of granularity; the throughput improved constantly while response time increased significantly; and at the same time, CFS was more suitable for different application environments by using different optimization strategies.