Minimal-delay decentralized maintenance of processor-group membership in TDMA-bus LAN systems

Abstract

Decentralized approaches to processor-group maintenance (GMM) are aimed at facilitating every active node in a real-time LAN system to maintain timely and consistent knowledge about the health status of all cooperating nodes and to recognize newly joining nodes. A practical scheme for this decentralized GMM (DGMM) in TDMA (time division multiplexed access) bus based real-time LAN systems was formulated earlier by Kopetz et al. The scheme is called here the periodic reception history broadcast (PRHB) scheme. This scheme was initially formulated for the application environments where the fault frequency is relatively low such that no more than one node fails in any interval of two TDMA cycle duration. In this paper, we develop a major extension of the scheme called the PRHB with multiple fault detection (PRHBMD) scheme applicable to the environments where the fault frequency is much higher, to be more specific, where up to a half of the nodes may experience faults within any interval of three TDMA cycle duration. The scheme does not impose any limit on the number of transient faults of links that any one node may experience. The scheme yields the minimal detection delay for all major fault types and the delay does not exceed two TDMA cycles for the worst fault type. This detection delay characteristics is a significant improvement over those of previously developed DGMM schemes.

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

“…However the central master in this approach constitutes is a single point of failure. A decentralized approach, called periodic reception history broadcast (PRHB), was proposed in [57] to resolve this problem. In PRHB, each node is bound to broadcast periodic reception history, gathered between the last two TDMA cycle.…”

A period-based group membership strategy for nodes of TDMA networks

“…Algorithms handle both permanent and transient faults, typically with restrictions on fault interarrival rates (Kim and Shokri, 1993). Group membership algorithms cannot compensate for loss of network connectivity or semantically incorrect data that is syntactically correct.…”

“…Consensus is guaranteed to be reached within two rounds after a fault has been identified (Pfeifer, 2000). If a fault occurs in the group, additional faults are not tolerated while nodes in the group have inconsistent views of membership, although better fault tolerance is possible for some faults if a slightly longer time is allowed (Kim and Shokri, 1993).…”

“…These group membership services protect against a variety of faults including processor faults, link faults, and noise on the communication bus (Kim and Shokri, 1993). The faults may be permanent or transient.…”