Abstract
Real time applications play an important role in today's scenario. Many real time critical applications such as space missions, nuclear reactors, defense systems, air-traffic control systems, etc. are highly dependent on their correct functioning as well as timing constraint. A real time system is considered to function correctly even there is fault in the system. All hard real time applications are highly dependent on its deadline. Missing the deadline in such systems can cause system failure. To avoid failure in these systems we use a well-known approach called check pointing. Fault tolerance is an important aspect in real-time applications. It is the technique used to achieve the required results in the presence of fault. Fault tolerance can be achieved by check pointing approach for transient fault. But the issue is how many checkpoints should be applied with minimum energy consumption. Fault tolerance and energy efficiency is one of the big issues in real time systems. The systems such as sensor node require energy to operate, but energy consumption should be less. Most of the real time systems working in such an environment where fault can occur at any time must be fault tolerated. In this paper we propose a new check pointing method to tolerate transient fault with minimum energy consumption. The proposed method finds optimal number of checkpoints with minimum energy consumption. The energy minimization is achieved by using a conventional Dynamic Voltage Scaling (DVS) technique used for energy management in real time systems.
