PFL: Proactive Fuzzy-logic-based AQM Algorithm for Best-effort Networks

AQM is an effective mechanism to provide early congestion control carried on the intermediate router. In this paper, we introduce a novel, autonomic, intelligent and stateless proactive fuzzy logic AQM (PFL) to deal with uncertain network events. PFL has two advantages. First, a simple and efficient non-linear fuzzy control law is built for PFL, which makes PFL capture network congestion status easily and tunes the packet-drop probability adaptively. Second, in order to make up for the deficiency of fixed fuzzy rules, a self-configuration technique based on steady-state packet loss ratio is introduced to regulate the output of PFL dynamically. Following an exhaustive simulation evaluation, we demonstrate that PFL is insensitive to dynamic traffic loads, various RTT, and also is robust to unresponsive traffic. Simulation results manifest that PFL scheme outperforms other well-known AQMs without the need to reset any parameters.