Design of a fault-tolerant COTS-based bus architecture

Abstract

In this paper, we report our experiences and findings on the design of a fault-tolerant bus architecture comprised of two COTS buses, the IEEE 1394 and the I2C. This fault-tolerant bus is the backbone system bus for the avionics architecture of the X2000 program at the Jet Propulsion Laboratory. COTS buses are attractive because of the availability of low cost commercial products. However, they are not specifically designed for highly reliable applications such as long-life deep-space missions. The X2000 design team has devised a multi-level fault tolerance approach to compensate for this shortcoming of COTS buses.First, the approach enhances the fault tolerance capabilities of the IEEE 1394 and I2C buses by adding a layer of fault handling hardware and software. Second, algorithms are developed to enable the IEEE 1394 and the I2C buses assist eacfi other to isolate and recovery from faults. Third, the set of IEEE 1394 and I2C buses is duplicated to further enhance system reliability. The X2000 design team has paid special attention to guarantee that all fault tolerance provisions will not cause the bus design to deviate from the commercial standard specifications. Otherwise, the economic attractiveness of using COTS will be diminished. The hardware and software design of the X2000 fault-tolerant bus are being implemented and flight hardware will be delivered to the ST4 and Europa Orbiter missions.

Survey and future directions of fault-tolerant distributed computing on board spacecraft

“…In [64,65] D. A. Rennels proposed a distributed hierarchical system, which extended the master/slave approach in [66]. The design is comprised of low-level and high-level computers.…”

“…Three types of approaches can be identified. In first type -AIPS [63] and FTD Computer [64,65] -multiple redundant processors are used to carry out a set of tasks in existing systems. In the second type -Mars [3], and Delta-4 [67] -hardware redundancy and software replication are employed.…”

Analysis of a multi-layer fault-tolerant COTS architecture for deep space missions

“…To compensate for COTS technologies' weakness in fault tolerance, the X2000 has employed a multi-level fault protection methodology to achieve high reliability [1] [2]. The methodology is applied to the X2000 bus architecture by using four levels of fault protection mechanisms.…”

“…To further enhance the effectiveness of the system level redundancy, diversity is also employed in this level. For example, the X2000 implements the redundant IEEE 1394 buses with different topologies, such that any branch node in primary bus set is a leaf node in the backup bus set and vice versa [1] [2]. In other words, there is no node that is a branch node for both buses.…”

“…In order to compensate for such fundamental weaknesses, different types of buses may be used to complement the IEEE 1394 bus. Specifically, the I 2 C bus, which has a multi-drop bus topology [5] [6], is used to assist the IEEE 1394 fault isolation and recovery in X2000 [1] [2]. The coordination between these two bus would require some custom software.…”