Paul M. Julich

Fault Tolerant Computer System for the A129 Helicopter

The A129 integrated multiplex system (IMS) is a highly reliable computer system designed to implement automatic flight control, navigation, system monitoring, and other flight-critical and mission related tasks. The reliability of the IMS has been achieved through the development of hardware-implemented and software implemented fault-tolerance techniques which exploit several unique architectural and hardware characteristics. This paper describes the fault-tolerance design philosophy, the IMS architecture, the fault detection and fault recovery techniques, and the hardware and software structures. Also presented is a Markov reliability analysis which was used to quantify the reliability of the system.

Distributed intelligent network management for the SDI ground network

The authors describe an integrated solution to distributed intelligent network management for the Strategic Defense System (SDS) terrestrial network which emphasizes the partitioning of the problem between rule-based, constraint-based, and algorithmic techniques. Problems in current approaches to network management for the SDS are described which show how this new approach leads to an efficient implementation in a C-based environment via the Harris-developed constraint propagation expert system shell and the C language integrated production system shell. The advantages of constraint-based distributed problem solving are presented. The authors describe how a model of battle management and the evolving battle situation can be used to perform proactive as well as reactive definition of routes, and other network functions.<<ETX>>