Is this project an undergraduate, graduate, or faculty project?
Undergraduate
group
What campus are you from?
Daytona Beach
Authors' Class Standing
Morgan Kendall, Sophomore
Lead Presenter's Name
Grace Gratton
Faculty Mentor Name
Dr. Bryan Watson
Abstract
Robustness to faulted agents in consensus algorithms within a multi-agent system is a critical requirement for reliable systems. Faulted agents can compromise the swarm, leading to catastrophic consequences, including economic impacts and possible loss of life. Therefore, it is critical to develop robust consensus algorithms to ensure the security and reliability of the systems we rely upon. Biologically Inspired Design previously led to the Synchronous Hatching Consensus Algorithm, which was robust up to 20% faulted agents reporting false positives, but not robust to faulted agents reporting false negatives. This work builds upon the previous model by increasing robustness to both false positives and false negatives, and by expanding model functionality to control the number of agents per nest, nest locations, and exit locations, to provide application insights. The model was tested against 0, 1, 5, 10, 15, 20, 30, and 40 faulted agents for false positives, negatives, and combined across four separate environments. Robustness to faulted agents was measured by reaching 66% consensus across all faulted agents and environments, with less than 5% error. Overall, grid nests with either even or random agents per nest performed optimally, with 0% failure and the fastest time to reach consensus, as compared to random nests with either even or random agents per nest. The model was also proven robust to up to 20% faulted agents for both false positives and false negatives.
Did this research project receive funding support from the Office of Undergraduate Research.
No
Evidence for Robustness of Synchronous Hatching Consensus Algorithm: Examining New Model Parameters and New Fault Types
Robustness to faulted agents in consensus algorithms within a multi-agent system is a critical requirement for reliable systems. Faulted agents can compromise the swarm, leading to catastrophic consequences, including economic impacts and possible loss of life. Therefore, it is critical to develop robust consensus algorithms to ensure the security and reliability of the systems we rely upon. Biologically Inspired Design previously led to the Synchronous Hatching Consensus Algorithm, which was robust up to 20% faulted agents reporting false positives, but not robust to faulted agents reporting false negatives. This work builds upon the previous model by increasing robustness to both false positives and false negatives, and by expanding model functionality to control the number of agents per nest, nest locations, and exit locations, to provide application insights. The model was tested against 0, 1, 5, 10, 15, 20, 30, and 40 faulted agents for false positives, negatives, and combined across four separate environments. Robustness to faulted agents was measured by reaching 66% consensus across all faulted agents and environments, with less than 5% error. Overall, grid nests with either even or random agents per nest performed optimally, with 0% failure and the fastest time to reach consensus, as compared to random nests with either even or random agents per nest. The model was also proven robust to up to 20% faulted agents for both false positives and false negatives.