We all hope for the best but sometimes, one must plan for ways of dealing with the worst-case scenarios, especially in a network with adversaries. This dissertation illustrates a detailed description of distributed optimization algorithms over a network of agents, in which some agents are adversarial. The model considered is such that adversarial agents act to subvert the objective of the network. The algorithms presented in this dissertation are solved via gradient-based distributed optimization algorithm and the effects of the adversarial agents on the convergence of the algorithm to the optimal solution are characterized. The analyses presented establish conditions under which the adversarial agents have enough information to obstruct convergence to the optimal solution by the non-adversarial agents. The adversarial agents act by using up network bandwidth, forcing the communication of the non-adversarial agents to be constrained. A distributed gradient-based optimization algorithm is explored in which the non-adversarial agents exchange quantized information with one another using fixed and adaptive quantization scheme. Additionally, convergence of the solution to a neighborhood of the optimal solution is proved in the communication-constrained environment amidst the presence of adversarial agents.
| Identifer | oai:union.ndltd.org:ucf.edu/oai:stars.library.ucf.edu:etd2020-2832 |
| Date | 15 August 2023 |
| Creators | Emiola, Iyanuoluwa |
| Publisher | STARS |
| Source Sets | University of Central Florida |
| Language | English |
| Detected Language | English |
| Type | text |
| Format | application/pdf |
| Source | Electronic Theses and Dissertations, 2020- |
Page generated in 0.0158 seconds