Hardware System Reliability Design And Analysis Of Satellite-carried Parallel Computer

This paper studies and analyzes the hardware system reliability design of satellite-carried parallel computer. We summarize the characteristics of high reliable satellite-carried parallel computer architecture design. The device reliability techniques are investigated from the points of space environments and radiation tolerance of satellite-carried computer. As the main functional parts of the satellite-carried computer, the reliability design of the space level microprocessor, memory and FPGA are thoroughly analyzed, combined with a discussion on the main failure mechanisms of satellite-carried microelectronic devices in space application. The on-chip fault-tolerance of space microprocessor is investigated. The application of redundancy, fault-tolerance techniques such as fault detection in satellite-carried computer are also covered in this paper. Fault-tolerance is the main technique of enhancing the reliability of satellite-carried computer. This paper studied the fault-tolerant design and realization of satellite-carried computer. We generalize the combinatorial reliability model for different architectures and compare the reliability of them, based on the introduction of reliability terms and analysis of computer system performance and dependability models. We further setup the Markov dependability model of hardware system of satellite-carried parallel computer and analyze the availability of this system. Finally, the performance of satellite-carried computer is studied.