2006 IEEE International Performance Computing and Communications Conference (2006)
Phoenix, AZ, USA
Apr. 10, 2006 to Apr. 12, 2006
DOI Bookmark: http://doi.ieeecomputersociety.org/10.1109/.2006.1629428
D.Y. Eun , Dept. of Electr.&Comput. Eng., North Carolina State Univ., Raleigh, NC, USA
X. Wang , Dept. of Electr.&Comput. Eng., North Carolina State Univ., Raleigh, NC, USA
For TCP/AQM systems, the issue of buffer sizing has recently received much attention. In the recent literature, it is suggested that the buffer size be O(/spl radic/N) for high link utilization contrasting to the traditional bandwidth-delay product, i.e., O(N) for 100% utilization where the link has capacity NC serving N flows. However, these results are all limited to the drop-tail scheme and there has been no systematic modeling framework for any buffer sizing between O(/spl radic/N) and O(N). In this paper, we study the limiting behavior of a TCP/AQM system for an intermediate buffer sizing of O(N/sup /spl gamma//) (0.5/spl les//spl gamma/<1). We develop a stochastic model in a discrete-time setting to characterize the system dynamics and then show that we can have 100% link utilization and zero packet loss probability for a large number of flows when the buffer size requirement is anywhere between O(/spl radic/N) and O(N). Our model is general enough to cover any queue-based AQM scheme with ECN marking (including the drop-tail) and various generalized AIMD algorithms for each TCP flow.
generalized AIMD algorithm, active queue management, TCP-AQM, buffer size, bandwidth-delay product, drop-tail scheme, stochastic model, probability, ECN marking
D. Eun and X. Wang, "Performance modeling of TCP/AQM with generalized AIMD under intermediate buffer sizes," 2006 IEEE International Performance Computing and Communications Conference(PCC), Phoenix, AZ, USA, 2006, pp. 50.