A Byzantine solution to early detect massive attacks

The quality and the timeliness of the detection of massive attacks significantly limit their great danger. In this paper, we describe an existing solution based on a centralized treatment of threat reports generated by probes deployed at the edges of a national Cyber-space. We also propose a more reliable architecture based on a consensus algorithm that solves the interactive consistency problem under the Byzantine assumptions. We prove the correctness of our algorithm and show its contribution to the early detection of massive attacks.