Feedback manipulation flooding attack: Feasibility evaluation and impact quantification on Stream Control Transmission Protocol

摘要

Stream Control Transmission Protocol (SCTP) is a general purpose and relatively new transport layer protocol with several unique features. This paper highlights the concept of feedback protocol and identifies SCTP as a protocol operating in a closed-loop feedback manner. We conduct an in-depth security analysis of closed-loop feedback operation of SCTP congestion control and present a new attack scenario called feedback manipulation flooding attack (FMFA). We show that standard SCTP senders can be remotely exploited for generation of powerful and sustained Denial-of-Service attack flood by tactically manipulating feedback messages. To ascertain the attack feasibility and its potential impact, we simulate the FMFA scenario using one of the globally well-accepted network simulators (ns2). We also compare the feedback manipulation flooding attack with conventional brute-force flooding attacks and identify some of its exclusive characteristics. Further, we implement the FMFA attack scenario in Linux kernel and present real-world experimental results over the Internet to validate our simulation results.