On Adversary Robust Consensus Protocols Through Joint-Agent Interactions

A generalized family of adversary robust consensus protocols is proposed and analyzed. These are distributed algorithms for multiagent systems seeking to agree on a common value shared variable, even in the presence faulty or malicious agents, which updating their local state according protocol rules. In particular, we adopt monotone joint-agent interactions, very general mechanism processing locally available information allowing cross-comparisons between state-values multiple agents simultaneously. The salient features class abstracted as Petri net convergence criteria resulting time evolutions formulated by employing structural invariants net.