A trust‐based resilient consensus algorithm for distributed optimization considering node and edge attacks
Summary Distributed optimization algorithms have various advantages and wide applications, while the distributed nature also makes them flimsy in the face of external attacks. In this paper two Byzantine attack models are designed, in which the edge attack model is seldom considered before. A resili...
Uloženo v:
| Vydáno v: | International journal of robust and nonlinear control Ročník 33; číslo 6; s. 3517 - 3534 |
|---|---|
| Hlavní autoři: | , |
| Médium: | Journal Article |
| Jazyk: | angličtina |
| Vydáno: |
Bognor Regis
Wiley Subscription Services, Inc
01.04.2023
|
| Témata: | |
| ISSN: | 1049-8923, 1099-1239 |
| On-line přístup: | Získat plný text |
| Tagy: |
Přidat tag
Žádné tagy, Buďte první, kdo vytvoří štítek k tomuto záznamu!
|
| Shrnutí: | Summary
Distributed optimization algorithms have various advantages and wide applications, while the distributed nature also makes them flimsy in the face of external attacks. In this paper two Byzantine attack models are designed, in which the edge attack model is seldom considered before. A resilient distributed consensus algorithm based on the trusted nodes and edges is proposed to defend against the two attack models. The agreement and optimality of the proposed consensus algorithm are guaranteed with theoretical analysis. The simulation results and comparison experiments verify the proposed algorithm to be more effective than the compared one. In the perspective of attackers, which attack model is more recommended is also analysed based on the experiments. |
|---|---|
| Bibliografie: | Funding information National Natural Science Foundation of China, Grant/Award Number: 62276062; Jiangsu Provincial Key Laboratory of Networked Collective Intelligence, Grant/Award Number: BM2017002 ObjectType-Article-1 SourceType-Scholarly Journals-1 ObjectType-Feature-2 content type line 14 |
| ISSN: | 1049-8923 1099-1239 |
| DOI: | 10.1002/rnc.6568 |