Detecting The Timeout Of Real-time Task Based On Dendritic Cell Algorithm

With the constant development of IT, especially the acceleration of combination in 3C, the requirement of real-time application is continuously expanding. Accordingly, embedded real-time operating system (RTOS) plays more and more important role. However, the correctness of RTOS not only depends on logic results of calculation, but also relies on the time of result's emergence. Timeout exception results in serious consequences. Then, how to timely and accurately find out timeout exception and take corresponding measures becomes an important problem. While artificial immune technology has been a hot research topic in the field of engineering application and intelligent computing, its function characteristics of resisting invasion open up a new research approach for computer security area.The paper aims to using artificial immune technology to solve real-time problems in RTOS. As traditional methods generally have shortcomings such as time-consuming, high cost and so on, it is proposed to introduce dendritic cell algorithm (DCA) into abnormal detection field, to early find out timeout problems appeared in real-time task sets, and to seek new approaches and methods to enhance reliability and stability of RTOS.Firstly, as task scheduling is a key factor of RTOS, the paper sums up the advantages and disadvantages of common scheduling strategies, and gives reasons of timeout and common solution. Besides, it analyses and points out shortages and defects when traditional technology detect timeout exception, and puts forward the timeout problems existed in real-time task sets.Secondly, the paper introduces biological mechanism of artificial immune system (AIS), new findings and its research status. Sale problems and self set changes existed in classic immune theory (SNS) are pointed out, DCA based on danger theory is expounded in detail. After deeply research on DCA, it is proposed to map time features of real-time task into algorithm parameters, signal and antigen. Besides, application framework that DCA detects timeout exception is designed, which refering to Libtissue - the classic framework of AIS.At last, through simplifying timeout problems, model of detecting timeout of real-time tasks is established. Based on the model, real-time tasks are simulated and scheduled, simultaneously, the simulation experiments that DCA detects timeout exception are finished. Through the comparison of DCA with traditional methods, the experiment verifies the effectiveness that DCA detects task timeout exception of RTOS (real-time task sets). Although the experiment does not provide solution to timeout exceptions, but the results indicate that DCA is more effective in indentifying the error reason than traditional methods. It is proved that DCA can accurately find the reason that cause timeout exceptions.