Effects of Ensemble-TCP. Eggert, L., Heidemann, J., & Touch, J. ACM Computer Communication Review, 30(1):15–29, January, 2000.
Effects of Ensemble-TCP [link]Paper  abstract   bibtex   
TCP currently recalculates the state of each connection from a fixed set of initial parameters; this recalculation occurs over several round trips, during which the con-nection can be less than efficient. TCP control block sharing is a technique for reusing information among connections in series and aggregating it among connec-tions in parallel. This paper explores the design space of a modified TCP stack that utilizes these two ideas, and one possible design (E-TCP) is presented in detail. E-TCP has been designed so that the network transmission behavior of group of parallel E-TCP connections closely resembles that of a single TCP/Reno connection. Simulated web accesses using HTTP/1.0 over E-TCP show a significant performance improvement compared to TCP/Reno connection bundles. This paper is first to evaluate performance using four different intra-ensem-ble schedulers for different workloads. In one scenario simulating a common case, E-TCP is 4-75% faster than Reno for transmitting the HTML parts of various pages, and 17-61% faster transmitting the whole pages. In the same scenario, reusing cached state speeds up repeated E-TCP page accesses by 17-53% for the HTML parts and 10-28% for the whole pages, compared to the initial access. E-TCP can also be integrated with other pro-posed TCP extensions (such as TCP/Vegas or TCP/SACK), to further improve performance.
@Article{Eggert00a,
	author = 	"Lars Eggert and John Heidemann and Joe Touch",
	title = 	"Effects of Ensemble-TCP",
	journal = 	"ACM Computer Communication Review",
	year = 		2000,
	sortdate = 		"2000-01-01",
	project = "ant, lsam",
	jsubject = "www",
	volume =	30,
	number =	1,
	month =		jan,
	pages =		"15--29",
	jlocation =	"johnh: folder: lsam",
	keywords =	"ensemble-tcp, tcp batching",
	url =		"https://ant.isi.edu/%7ejohnh/PAPERS/Eggert00a.html",
	psurl =		"https://ant.isi.edu/%7ejohnh/PAPERS/Eggert00a.ps.gz",
	pdfurl =		"https://ant.isi.edu/%7ejohnh/PAPERS/Eggert00a.pdf",
	myorganization =	"USC/Information Sciences Institute",
	copyrightholder = "author",
	abstract = "
TCP currently recalculates the state of each connection from a fixed
set of initial parameters; this recalculation occurs over several
round trips, during which the con-nection can be less than
efficient. TCP control block sharing is a technique for reusing
information among connections in series and aggregating it among
connec-tions in parallel. This paper explores the design space of a
modified TCP stack that utilizes these two ideas, and one possible
design (E-TCP) is presented in detail. E-TCP has been designed so that
the network transmission behavior of group of parallel E-TCP
connections closely resembles that of a single TCP/Reno
connection. Simulated web accesses using HTTP/1.0 over E-TCP show a
significant performance improvement compared to TCP/Reno connection
bundles. This paper is first to evaluate performance using four
different intra-ensem-ble schedulers for different workloads. In one
scenario simulating a common case, E-TCP is 4-75\% faster than Reno for
transmitting the HTML parts of various pages, and 17-61\% faster
transmitting the whole pages. In the same scenario, reusing cached
state speeds up repeated E-TCP page accesses by 17-53\% for the HTML
parts and 10-28\% for the whole pages, compared to the initial
access. E-TCP can also be integrated with other pro-posed TCP
extensions (such as TCP/Vegas or TCP/SACK), to further improve
performance.
",
}

Downloads: 0