The Study Of Real-Time Systems’ Survivability Based On Spn

With the popularity of social modernization and expansion of computer applications, real-time systems play an increasingly important role in our daily lives. Although traditional security measures can be used to protect the security of real-time systems, the accidents still exist. The capacity of providing basic services after an unexpected accident can be called survivability, and survivability is more important than security in the performance of a real-time system. Therefore, real-time system has become a measure of quality of service indicators. Resulting in higher survivability, the system is still able to provide basic services in the event of failure or accidents,Currently, survivability security is a hotspot in the field of the real-time system, which can be used as an indicator to measure the quality of services in a variety of real-time systems. Stochastic Petri nets (Stochastic Petri Net, SPN) have a strong dynamic analysis capability for asynchronous systems, concurrency and uncertainty which can be used to model the system survivability and analysis and evaluation.The main contributes of this thesis are based on the theory of real-time systems SPN survivability research. Firstly, SPN model real-time systems, and then gives the SPN model related Markov chain. Finally the state transition matrix is obtained by the Markov chain, which gives us the survivability parameters.Firstly, the real-time system survivability are analyzed and studied in the thesis, and the factors that affect single real-time system survivability are identified; Secondly, based on the evaluation factors set single factor, SPN model of evaluation criteria is established; and then the actual SPN model is built according to the random data. Finally, the similarity between two models is compared, and the most similar model SPN is identified. Around the SPN model theory and research, on the basis of the model, a variety of survivability parameters are given, and the simulation is conducted. Based on SPN model, this thesis presents a viable assessment method. Therefore, the SPN, introduced into the field of survivability, will greatly improve capabilities of the real-time system survivability.