Sponsor
This research was supported in part by DARPA contracts/grants N66001-97-C-8522 and N66001-97-C-8523, by NSF grants CCR-9876217 and by Tektronix Inc., and Intel Corporation.
Published In
Proceedings Japan-USA-Vietnam Workshop on Research and Education in Systems, Computation and Control Engineering
Document Type
Conference Proceeding
Publication Date
6-2000
Subjects
Computer network protocols -- Technological innovations, Computer networks, TCP/IP (Computer network protocol)
Abstract
Experimentally, we show that Transmission Control Protocol (TCP)’s congestion control algorithm results in dynamic behavior similar to a stable limit cycle (attractor) when data from TCP flow into a fixed-size buffer and data is removed from the buffer at a fixed service rate. This setup represents how TCP buffers packets for transmission onto the network, with the network represented by a fixed-size buffer with a fixed service rate. The closed trajectory may vary slightly from period to period due to the discrete nature of computer systems. The size of the closed trajectory is a function of the network’s buffer size and the network’s service rate of packets in the buffer.
Persistent Identifier
http://archives.pdx.edu/ds/psu/10394
Citation Details
"Application of Control Theory to Modeling and Analysis of Computer Systems," Molly H. Shor, Kang Li, Jonathan Walpole, David Steere, and Calton Pu, In proceedings Japan-USA-Vietnam Workshop on Research and Education in Systems, Computation and Control Engineering, HoChiMinh City, Vietnam, June 7-9, 2000.
Included in
Computer and Systems Architecture Commons, Digital Communications and Networking Commons