A systematic fault-tolerant computational model for both crash failures and silent data corruption

Cui, Xiaopeng; Hussain, Zaeem; Znati, Taieb; Melhem, Rami

doi:10.1109/icin.2018.8401596

Cited by 1 publication

(1 citation statement)

References 22 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…Fault tolerance schemes have been designed for modeling crash failures and silent data corruption within cloud computing systems using a parallel computing model [12]- [15]. The failure mechanisms have been modeled through the construction of an Optimisation problem.…”

Section: Introductionmentioning

confidence: 99%

Hybrid models for computing fault tolerance of IoT networks

Chokara

Jammalamadaka

2023

TELKOMNIKA

View full text Add to dashboard Cite

Many Internet of Things (IoT) -based networks are being built to develop applications spanning multiple domains. Many small to large devices connected in various ways increases the risk of IoT networks failing. Small devices in the devices layer frequently fail due to their small size and high usage. Intermittent failures of the IoT networks lead to catastrophes at times. The IoT systems must be designed to be fault-tolerant. Fault tolerance of IoT networks must be computable so that the same can be considered while designing IoT networks. However, the computation of fault tolerance of IoT networks is complex, especially when heterogeneous structures are used for building a specific IoT network. Fault tree-based models are not suitable for computing fault-tolerance of complex models, which requires probability assessment. Hybrid fault tolerance computing models have been presented in this paper that consider both linear and probabilistic methods of computing the fault tolerance considering many complex networking topologies used in each layer of IoT networks. The fault-tolerance computing models are formal methods that can be used to compute the fault tolerance of any IoT network built with any internal processing. The accuracy of fault tolerance computing is 12.9% higher than other methods.

show abstract

Section: Introductionmentioning

confidence: 99%