Adaptive middleware design for satellite fault-tolerant distributed computing

This paper is concerned with increasing the reliability with respect to debris impact of computing systems on board satellites using Fault Tolerant Distributed Computing. A novel adaptive middleware for fault-tolerance (AMFT) design is presented. The middleware performs failure detection, isolation and recovery and synchronizes the operation of a distributed On-Board Computer (OBC) system. If one of the members of the distributed group of OBCs fails, the failed processor tasks are reallocated dynamically to the other healthy OBC units. The approach can also be employed in other space applications such as launch vehicles, interplanetary rovers, etc. The proposed design is modular, low cost and modifiable according to the requirements of the target application.