Applying Neural Network to Improve TCP Performance over Distributed Spaceflight TT&C Networks

Author

Gong, Changqing ; Meng, Qingjie ; Zhao, Linna ; Bi, Xiaoxia

Author_Institution

Shenyang Inst. of Aeronaut. Eng., Shenyang

fYear

2007

fDate

23-25 May 2007

Firstpage

977

Lastpage

979

Abstract

A backpropagation neural network classifier is presented, which can classify packet loss cause over TT&C (tracking, telemetry, and command) networks. In TT&C networks, there are many packet losses are due to link errors, but not due to network congestions; and then TCP often reduces its data rate mistakenly, cannot keep a reasonable data rate. Based on the packet loss classifier, TCP-BP algorithm is proposed. The result of our simulation shows that the TCP-BP algorithm is superior to Vegas and Reno algorithm.

Keywords

backpropagation; neural nets; packet radio networks; space communication links; telemetry; transport protocols; TCP performance; TT&C networks; backpropagation neural network classifier; command networks; distributed spaceflight; link errors; packet loss classifier; telemetry networks; tracking networks; Aerospace engineering; Backpropagation algorithms; Computer networks; Distributed computing; Educational institutions; Input variables; Neural networks; Protocols; Satellites; Vibration measurement;

fLanguage

English

Publisher

ieee

Conference_Titel

Industrial Electronics and Applications, 2007. ICIEA 2007. 2nd IEEE Conference on

Conference_Location

Harbin

Print_ISBN

978-1-4244-0737-8

Electronic_ISBN

978-1-4244-0737-8

Type

conf

DOI

10.1109/ICIEA.2007.4318553

Filename

4318553