Dual-domain Defenses for Byzantine-resilient Decentralized Resource Allocation

TL;DR

This paper develops dual-domain defense algorithms for decentralized resource allocation that are resilient to Byzantine attacks, ensuring convergence to near-optimal solutions despite malicious agents.

Contribution

It introduces a novel class of Byzantine-resilient algorithms using dual-variable filtering and robust aggregation, advancing secure decentralized resource management.

Findings

Algorithms converge to neighborhoods of optimal allocation

Robust aggregation rules ensure Byzantine resilience

Numerical experiments validate theoretical guarantees

Abstract

This paper investigates the problem of decentralized resource allocation in the presence of Byzantine attacks. Such attacks occur when an unknown number of malicious agents send random or carefully crafted messages to their neighbors, aiming to prevent the honest agents from reaching the optimal resource allocation strategy. We characterize these malicious behaviors with the classical Byzantine attacks model, and propose a class of Byzantine-resilient decentralized resource allocation algorithms augmented with dual-domain defenses. The honest agents receive messages containing the (possibly malicious) dual variables from their neighbors at each iteration, and filter these messages with robust aggregation rules. Theoretically, we prove that the proposed algorithms can converge to neighborhoods of the optimal resource allocation strategy, given that the robust aggregation rules are properly designed. Numerical experiments are conducted to corroborate the theoretical results.