A Resilient Convex Mixture for consensus-based distributed algorithms
Authors: Xuan Wang, Shaoshuai Mou, Shreyas Sundaram
Summary: Take into account a set of vectors in Rn, partitioned into two courses: regular vectors and malicious vectors. The variety of malicious vectors is bounded however their identities are unknown. The paper offers a manner for reaching a resilient convex mixture, which is a convex mixture of solely regular vectors. In contrast with current approaches based mostly on Tverberg factors, the proposed technique based mostly on the intersection of convex hulls has decrease computational complexity. Simulations counsel that the proposed technique will be utilized to resilience for consensus-based distributed algorithms towards Byzantine assaults.
