A Comparative Study among Four Controllers Intended for Congestion Control in Computer Networks
Computer networks efficiency is an vital part of today’s information services technology, with this comes multiple issues, among them is the congestion problem. This paper will discuss the designing and evaluating of four controllers to deal with this issue. The design starts with modeling the Transmission Control Protocol /Active Queue Management (TCP/AQM) which is intended for dynamics modeling of the average TCP window size and the queue size in the bottleneck router. Apart from modeling, the work comprises of two parts. In the first, three controllers Random Early Detection, Proportional-Integral and Proportional-Integral-Derivative (RED, PI, and PID) are designed, tested, evaluated, and compared among each other, with the use of the TCP/AQM model developed. The second part considers designing a fuzzy logic based online tuned PID controller and comparing its performance with a PID controller tuned offline with three tuning methods, Ziegler Nichols (Z-N), Particle Swarm Optimization (PSO), and Ant Colony Optimization (ACO). The Integral Square Error (ISE) is used as the objective function for optimization. The controllers’ performance is evaluated using the following parameters for system’s response, rise time, settling time, and maximum peak overshoot. The performance of the controllers is also examined by applying a disturbance as an exceptional condition. To test and evaluate the controllers, the system as all is implemented using MatLab (Version 2014). The results obtained indicated that the PID gave a better performance, compared to the RED and the PI, in following changes in the desired queue level, and in reducing the loss of packets. The PID gave a settling time 20% lesser than that of the PI and 60% lesser than that of the RED. Regarding the tuning methods, and under the settings considered for each in this work, the ACO-PID gave the least overshoot (1.545%) compared to the others methods [ZN-PID (40%), PSO-PID (13.85%), Fuzzy-PID (5%)].
The PSO and ACO managed to cause great reduction in settling time () and rise time (). The ratios of and of PSO-PID to PID before tuning are (16.5%), (23.43%) and the ratio of and of ACO-PID to PID before tuning are (11.5%), (44.56%). The intelligent tuning methods [PSO & ACO] gave better and compared to Fuzzy or Ziegler–Nichols. Despite the indicated relative performance of the Fuzzy PID controller, it has some important privileges. Firstly, it is an online tuning method, as it continuously adapts the PID controllers’ parameters as long as the system is running. Secondly, its performance can still be improved by optimizing the fuzzy part. Thirdly, it represents a nonlinear controller (as its parameters are changing), and so it can even suit the nonlinear model.
V. Jacobson and M. J. Karels. “Congestion Avoidance and Control.” in Proc.ACM SIGCOMM, 1988, pp. 314-329.
G. F. Ahammed and R. Banu. “Analyzing the Performance of Active Queue Management Algorithms.” in (IJCNC) International Journal of Computer Networks & Communications, Vol.2, No.2, March 2010.
T. Azuma, T. Fujita, and M. Fujita, “Congestion control for TCP/AQM networks using state predictive control.” EEE Transactions on Electronics, Information and Systems, Vol. 125, No.9, pp. 1491–1496, 2005.
C. V. Hollot, V. Misra, D. Towsley and W. Gong, “Analysis and Design of Controllers for AQM Routers Supporting TCP flows.” IEEE Transactions on Automatic Control, Vol. 47, No. 6, pp. 945-959, 2002.
V. Misra, W. B. Gong and D. Towsley, “ Fluid-Based Analysis of a Network of AQM Routers Supporting TCP Flows with an Application to RED.” in Proc. ACM Sigcomm, 2000, pp. 151-160.
C. V. Hollot, V. Misra, D. Towsley, and W. B. Gong, “ A Control Theoretic Analysis of RED.” in Proceedings of IEEE INFOCOM, 2001.
C. V. Hollot, V. Misra, D. Towsley, and W. B. Gong, “ On Designing Improved Controllers for AQM Routers Supporting TCP Flows.” in Proceedings of IEEE INFOCOM, 2001, pp. 1726–1734.
M. Y. Waskasi, M. J. Yazdanpanah and N. Yazdani, “ a New Active Queue Management Algorithm Based On Neural Networks PI.” Control and Intelligent Processing Center of Excellence Electrical and Computer Engineering Department, University of Tehran, 2005
M. Z. Al-Faiz and A. M. Mahmood, “ Fuzzy-Genetic Controller for Congestion Avoidance in Computer Networks.” IJCCCE, Vol. 11, No. 2, 2011.
T. Alvarez, “ Design of PID Controllers for TCP/AQM Wireless Networks.” in Proc. the World Congress on Engineering, 2012, pp. 1273-1280.
S. T. Salim and A. M. Mahmood, “ Design of On-Line Tuned Controller for Congestion Avoidance in Computer Networks.” Eng. &Tech. Journal, Vol. 31, No. 3, 2013.
H. I. Ali, and K. S. Khalid, “ Swarm intelligence based robust active queue management design for congestion control in TCP network.” IEEJ Transactions on Electrical and Electronic Engineering, Vol. 11, pp. 308-324, 2016.
J. Olsen, “ Stochastic modeling and simulation of the TCP protocol.” Uppsala Dissertations in Mathematics, Uppsala University, Sweden, 2003.
Ch. Koutsimanis, and P. G. Park, “ Active Queue Management – A router based control mechanism.” Scientific report, 2006.
V. Chopra, S. K. Singla, and L. Dewan, “Comparative Analysis of Tuning a PID Controller using Intelligent Methods,” Acta Polytechnica Hungarica, Vol. 11, No. 8, 236-249, 2014
B. R. Copeland, “The Design of PID Controllers using Ziegler Nichols Tuning,” Internet: http://educypedia.karadimov.info/library/Ziegler_Nichols.pdf, Mar. 2008[ Apr. 20017].
B. Nagaraj and N. Murugananth, “A comparative study of PID controller tuning using GA, EP, PSO and ACO,” Journal of Automation, Mobile Robotics & Intelligent Systems, Vol. 5, No. 2, 2011.
J. Jin and H. Huang, “Study on fuzzy self-adaptive PID control system of biomass boiler drum water,” Journal of bio-energy systems, Vol. 3, No. 1, 2013.
- There are currently no refbacks.