Title :
A Congestion Control Approach Based on Predictive Model in Delay Network
Author :
Xiao, Ping-Ping ; Tian, Yan-tao
Author_Institution :
Sch. of Inf. Sci. & Eng., Jilin Teacher´´s Inst. of Technol. & Eng., Changchun
Abstract :
To overcome the disadvantage influence of time delay term on the TCP/AQM control system model, the paper proposed that a new AQM algorithm called SPI (proportional integral control based on Smith predictor), which combined Smith predictor and Dahlin algorithm. By moving the time delay term to the outside of control loop, SPI can eliminate the influence of time delay on the system performance, and reduce the number of the tuning parameters. The result of simulation shows that the presented SPI controller can effectively avoid network congestion, improve the system stability, and have a good performance in respect to robustness to variation of system parameters
Keywords :
PI control; delays; queueing theory; telecommunication congestion control; AQM algorithm; Dahlin algorithm; PI controller; Smith predictor model; TCP control system model; active queue management; congestion control approach; delay network; proportional integral control; time delay; Control system synthesis; Control systems; Delay effects; Pi control; Prediction algorithms; Predictive models; Proportional control; Robust control; Robust stability; System performance; Communication technology; PI controller; active queue management; congestion control;
Conference_Titel :
Machine Learning and Cybernetics, 2006 International Conference on
Conference_Location :
Dalian, China
Print_ISBN :
1-4244-0061-9
DOI :
10.1109/ICMLC.2006.259159
