International Journal of Computer Science and Technology

ISSN 2996-8223

International Journal of Computer Science and Technology | Vol. 6, No. 2, February 2004 | pp. 9–16

DOI: 10.46882/2004/IJCST/000055

Article Type: Research Article

Title: An Adaptive Fuzzy Logic Controller for Dynamic Buffer Management in Core Internet Routers

Names of Authors: Ming-Wei Chen¹, Sarah Jenkins²

Authors’ Affiliations: ¹Department of Software Engineering, Tsinghua University, Beijing, China; ²Department of Cybersecurity, MIT, Massachusetts, USA

Abstract: Traditional active queue management techniques, such as Random Early Detection (RED), struggle to maintain low queuing delays under heavily fluctuating Internet traffic workloads. This study proposes an Adaptive Fuzzy Queue Controller (AFQC) designed to optimize packet drop probabilities in core router buffers. The controller utilizes a non-linear two-input fuzzy inference engine that maps both instantaneous queue length variations and traffic arrival acceleration metrics. Network simulations conducted via NS-2 utilizing heavy-tailed Pareto traffic profiles showed that AFQC stabilizes buffer occupancy levels within a narrow ± 8.5% margin of the target threshold. Compared to standard RED implementations, the proposed algorithm reduced packet loss rates by 14.3% and limited average end-to-end packet jitter to under 3.2 ms. The processing overhead of the fuzzy membership evaluations takes less than 8 μs per packet, making it viable for high-speed line card deployment.

Keywords: Active Queue Management, Fuzzy Logic Control, Computer Networks, Router Architecture, Traffic Engineering

Manuscript Timeline: Received: November 02, 2003; Revised: December 14, 2003; Accepted: January 11, 2004; Published: February 12, 2004

Citation: Chen, M.-W., & Jenkins, S. (2004). An Adaptive Fuzzy Logic Controller for Dynamic Buffer Management in Core Internet Routers. International Journal of Computer Science and Technology, 6(2), 9–16.