Mean Waiting Delay for Web Object Transfer
in Wireless SCTP Environment
Abstract-Most current web application use HTTP (Hyper Text
Transfer Protocol) and TCP (Transmission Control Protocol) to
retrieve objects from the Internet. SCTP (Stream Control
Transmission Protocol) is recently proposed transport protocol
with congestion control mechanism similar to that of TCP.
Waiting delay is an important performance criterion for when
transferring web object over the Internet. In this paper, we
present an analytical model of mean waiting delay for object
transfers over the Internet in a wireless using the SCTP as the
transport protocol and compare with the mean waiting in using
TCP. Validation of the model using experimental results show
that the mean waiting delay for HTTP over SCTP is less than
that for HTTP over TCP. This is caused by the small slow-start
time of SCTP.
I. INTRODUCTION
Most web applications use HTTP (hyper text transfer
protocol) as the transfer protocol to retrieve objects in the
Internet. HTTP is a connection-oriented protocol and uses
TCP (transmission control protocol) as the transport layer
protocol. TCP provides a single stream of data and strict
ordered delivery. Consequently, when a packet is lost in the
network, all subsequent packets arriving at the receiver must
wait until the lost packet has been retransmitted from the
sender and received at the receiver. This phenomenon, called
Head of Line (HOL) blocking, results in waiting delay [1] that
affects the performance of web transfers.
SCTP (Stream Control Transmission Protocol) [2,3,4] has
been proposed by IETF as a new transport layer protocol. The
design of SCTP absorbed many strengths of TCP, such as
window-based congestion control, error detection, and
retransmission. Moreover, SCTP incorporated several new
features that are not available in TCP. Two main new features
are multi-streaming and multi-homing. SCTP’s multi-
streaming can alleviate the head-of-line blocking when an
HTML page contains multiple objects. However, it cannot
avoid the waiting delay between packets when single objects
are transferred; the end point must hold the received packet
from delivery to the upper layer protocol (HTTP) until
they are reordered. This waiting delay is affected by slow-start
and retransmission policy.
Several models [5,6,7] that estimate the mean transfer times
of TCP application have been presented in the literature;
however, they assumed single packet loss. Furthermore, since
the waiting delay was not considered in the transfer time, we
cannot extract the waiting delay from the previous models.
The objective of this paper is to develop an analytical model
to estimate the waiting delay for multiple packet losses when
SCTP is used in a wireless web environment and compare
with the mean waiting delay of TCP.
The rest of the paper is organized as follows. Section 2
provides background on the web object transfer in TCP and
SCTP environment. Section 3 describes the mean waiting
delay time in web object transfer. Section 4 presents mean
waiting delay comparison of HTTP over TCP and HTTP over
SCTP. Finally, concluding remarks are given in Section 5.
II. WEB OBJECT TRANSFER IN TCP AND SCTP
ENVIRONMENT
Consider the simple case of a web browser displaying a
web page with embedded objects. Using HTTP/1.1 that
supports persistent and pipelined connections, the browser
opens a new transport connection to the server, and sends an
HTTP GET request with the desired URI (Uniform Resource
Identifier). The request may consist of a specific command, a
message containing request parameters, information about the
client and may contain URI’s of embedded objects. The server
returns an HTTP response with the page contents. The
response includes status information, a success/error code,
and a message containing information about the server and
information about the response itself. As responses arrive
from the server, the browser displays the web page with its
embedded objects. In general, objects embedded within a web
page are independent of each other. That is, requesting and
displaying each object in the page does not depend on the
reception of other embedded objects. This leads to HOL
blocking problem that causes performance degradation.
To alleviate HOL blocking, web browsers usually open
multiple TCP connections to the same web server. Using
multiple TCP connections for transferring a single
application's data introduces many negative consequences for
both the application and the network. It may cause the
increased load on web server. Under high loads, some web
servers may choose to drop incoming TCP connection
requests due to lack of available memory resources. The other
drawback is increased connection establishment latency. Each
TCP connection goes through a three-way handshake for
Yong-Jin Lee
1
and M.Atiquzzaman
2
1
Department of Technology Education, Korea National University of Education
2
School of Computer Science, University of Oklahoma
This full text paper was peer reviewed at the direction of IEEE Communications Society subject matter experts for publication in the IEEE ICC 2009 proceedings
978-1-4244-3435-0/09/$25.00 ©2009 IEEE