Tolerância a falhas através de escalonamento em um sistema multiprocessado

Fault tolerance must be included in the initial design of real time systems, must encompass hardware and software, and must be integrated with timing constraints. In many situations, the fault-tolerant design must be static, due to extremely high data rates and severe timing constraints. Ultrareliable systems need to employ proof-of-correctness techniques to ensure fault tolerance properties. We also see new approaches where the system predicts that timing constraints will be missed, enabling early action on such faults. Prioritizing tasks in HardReal-Time Systems is a problem belonging to NP-hard class. Various alternatives for solving that problem have already been proposed. In the present study, we have investigated the possibility of obtaining a high level of fault-tolerance with the integration of some works proposals. This work presents techniques to enhance the fault-tolerance capability of hard real-time systems by incorporating time redundancy, processor redundancy and fault tolerant real-time communication protocols. The main goal is to guarantee fault-tolerance requisites for multiprocessor hard real-time systems. The resources used to optimize these faulttolerance requisites are task schedulability and time redundancy. This work, aside from theoretical concerns, will research the implementation of using a DSP multiprocessor system interconnected by a CAN bus. The research will consider the overhead introduced by the real-time fault tolerant communication protocol.