A Lightweight Strategy for Reliability of Consensus Mechanisms based on Software Defined Networks
| Autor: | Gabriel R. Carrara, Diogo M. F. Mattos, Célio Albuquerque, Lucio H. A. Reis |
|---|---|
| Rok vydání: | 2019 |
| Předmět: |
Computer science
business.industry Quality of service Reliability (computer networking) media_common.quotation_subject 05 social sciences Control (management) 050801 communication & media studies 020206 networking & telecommunications 02 engineering and technology Flow network Reduction (complexity) 0508 media and communications Voting 0202 electrical engineering electronic engineering information engineering business Software-defined networking Computer network media_common |
| Zdroj: | GIIS |
| DOI: | 10.1109/giis48668.2019.9044954 |
| Popis: | Private blockchain applications simplify the network consensus while relying on voting-based consensus mechanisms because these mechanisms present lower computational cost than proof-of-work. The voting-based consensus is adequate to networks where all nodes are known, although it requires a high amount of exchanged messages, proportional to the number of participant nodes. In this paper, we propose a strategy for applying Quality of Service (QoS) techniques to ensure the reliable operation of voting-based consensus mechanisms. The proposed strategy is based on Software Defined Networking (SDN) and aims at reducing the time for the termination of the consensus protocols. The proposal evaluation leans on an emulated environment, using Mininet, where we run the Raft and BFT-SMaRt consensus mechanisms. Results show that the proposal ensures the termination of the protocols in a reduced time and the reduction of the control load for the election of leaders since we assure a minimum quality of service on the throughput rate for the consensus network flows. The proposed strategy enhances up to 100% of the consensus terminations on Raft and, for BFT-SMaRt, ensured the operation even in contention scenarios where the consensus was previously unfeasible. Results also show 80% reduction in the exchanged messages for achieving consensus. |
| Databáze: | OpenAIRE |
| Externí odkaz: |
|