Design and simulation an optimal enhanced PI controller for congestion avoidance in TCP/AQM system

TELKOMNIKA Telecommunication Computing Electronics and Control
Vol. 21, No. 5, October 2023, pp. 997~1004
ISSN: 1693-6930, DOI: 10.12928/TELKOMNIKA.v21i5.24872  997
Journal homepage: http://telkomnika.uad.ac.id
Design and simulation an optimal enhanced PI controller for
congestion avoidance in TCP/AQM system
Yousra Abd Mohammed1
, Layla H. Abood2
, Nahida Naji Kadhim2
1
Communication Engineering Department, University of Technology, Baghdad, Iraq
2
Department of Control and System Engineering, University of Technology, Baghdad, Iraq
Article Info ABSTRACT
Article history:
Received Dec 05, 2022
Revised Mar 28, 2023
Accepted Apr 30, 2023
In this paper, snake optimization algorithm (SOA) is used to find the optimal
gains of an enhanced controller for controlling congestion problem in
computer networks. M-file and Simulink platform is adopted to evaluate the
response of the active queue management (AQM) system, a comparison with
two classical controllers is done, all tuned gains of controllers are obtained
using SOA method and the fitness function chose to monitor the system
performance is the integral time absolute error (ITAE). Transient analysis and
robust analysis is used to show the proposed controller performance, two
robustness tests are applied to the AQM system, one is done by varying the size
of queue value in different period and the other test is done by changing the
number of transmission control protocol (TCP) sessions with a value of ± 20%
from its original value. The simulation results reflect a stable and robust
behavior and best performance is appeared clearly to achieve the desired
queue size without any noise or any transmission problems.
Keywords:
AQM
Computer network
Congestion control
NLPI controller
SOA
This is an open access article under the CC BY-SA license.
Corresponding Author:
Yousra Abd Mohammed
Communication Engineering Department, University of Technology
Baghdad, Iraq
Email: yousra.a.mohammed@uotechnology.edu.iq
1. INTRODUCTION
Avoiding high rates of packet loss in the internet is an important issue, when a packet is losing before
it received from its destination point; all the data it has sent through transmission path are dissipated.
In excessive states, this case will cause a congestion collapse. Congestion control has become as an urgent
matter in computer and communication networks. Congestion has harmful effect on network effeciency which
reflecting to noticeable packet loss, poor utilization, high delay rate, few throughputs [1], [2].
Different algorithms is used to solve the congestion problem but the best one is the active queue
management (AQM) scheme, AQM algorithm maintains congestion by previously detection of incipient
congestion and giving feedback signals to end-hosts to permit them minimizing their transmission rate before the
router’s buffer exceeded [3]. In [4] an inclusive study is demonstrated on the AQM algorithm techniques that
suggested and adopted to modify the performance efficiently, the AQM algorithms are classified based on length
of the queue or its delay or both of them. Proportional-integral-derivative (PID) fuzzy-neural controller was
presented; PID controller gains were tuned based on particle swarm optimization (PSO) tuning algorith to enhance
the fuzzy controller behavior [5]. An AQM scheme is controlled using fuzzy controller as a (mixing-fuzzy-PID)
controller to provie best congestion solution and few delay periods, high utilization and few data drop [6]. A fuzzy
proportional integral (FPI) controller with genetic tunning was suggested as an AQM for internet router [7].
An enhanced version of discrete linear quadratic optimal controller is adapted to track the desired queue size
in AQM scheme and genetic algorithm (GA) are used to fix the complexity of finding the weighting matrices
𝑄 and 𝑅 [8].

 ISSN: 1693-6930
TELKOMNIKA Telecommun Comput El Control, Vol. 21, No. 5, October 2023: 997-1004
998
The adaptive suggested method was introduced as a newly adopted method, named PHAQM, by using the
Hebb neural network for adjusting the variables of the model predictive control (MPC)-based AQM system [9].
Different methods were formed to maintain the congestion problem at the router in the network [10]–[13].
The algorithm that used for congestion problem is the drop tail (DT) method [14]. This method adopted the
first-in first-out (FIFO) technique. DT method are used without thresholds, then if the capacity is maximized
at the router buffer, the packets that arrived are directly. Maintaining the problem of congestion in AQM system
based on a smart and unique snake optimization algorithm (SOA) tuning method for finding gains of the
controller to fulfill the stability and robustness by fixing any problem mdropped [12], [15]. In [16] a fuzzy
controlling method was used to enhance the modify AQM system behavior without need for a precise model.
In this paper an enhanced proportional integral (PI) controller is proposed for solve congestion
problem that occurs in the communication networks during transmission data. The rest of the paper is: part 2
explains the modelling of AQM system. Part 3 indicateds the controller suggested scheme. Part 4, demonstrates
the SOA tuning algorithm, part 5 discuss the results obtained then part 6 presents the paper conclusions.
2. MODELING OF TCP/AQM
Transmission control protocol (TCP) operation was modeled and studied using differential equations,
together with a fluid-flow-based technique for ignoring TCP timeout and an investigation of stochastic
relationships [17], [18]. It is assumed that each homogenous TCP flow connected to a certain bottleneck
topology has an identical delay, as shown in Figure 1 [19]. The dynamic model is defined by the nonlinear
deferential equations [20], [21] shown below, where the parameters of (1), (2) and (3) are shown in Table 1.
𝑊(𝑡)
̇ =
1
𝑅(𝑡)
−
𝑊(𝑡)𝑊(𝑡−𝑅(𝑡))
2𝑅(𝑡−𝑅(𝑡))
𝑃(𝑡 − 𝑅(𝑡)) (1)
𝑞(𝑡)
̇ =
𝑊(𝑡)
𝑅(𝑡)
𝑁(𝑡) − 𝐶 (2)
𝑅 can be determined from (2) as shown in (3).
𝑅 =
𝑞
𝑐
+ 𝑇𝑝 (3)
Figure 1. Bottleneck scenario [19]
Table 1. Parameters of AQM model [21]
Symbol Description
(𝑊) TCP window size (packets) is positive, 𝑊 𝜖 [0, ˉ𝑊], ˉ𝑊: (max. size of window)
Ẃ(𝑡) Derivative time of 𝑊(𝑡)
(𝑞) Queue size (packet), 𝑞 𝜖 [0, −𝑞], ˉ𝑞 (buffer capacity)
𝑞́(𝑡) 𝑞(𝑡) derivative time
(𝑡) Time (sec)
(𝑅) Time of round trip (sec)
(𝑁) Number of TCP sessions
(𝐶) Capacity of the link (packet/sec)
(𝑝) Packet probability (mark/drop) [0, 1]
(𝑇𝑝) Propagation delay

TELKOMNIKA Telecommun Comput El Control 
Design and simulation an optimal enhanced PI controller for … (Yousra Abd Mohammed)
999
The nonlinear relations of the AQM model were linearized [8], [22] as shown in (4), Figure 2 shows
the linearized AQM control system.
𝑝(𝑠) = (
𝐶2
2𝑁
𝑒−𝑠𝑅0) ((𝑠 +
2𝑁
𝑅0
2𝐶
) (𝑠 +
1
𝑅0
))
⁄ (4)
Figure 2. Linearized AQM system block diagram [8]
In this study the values of the parameter used are: 𝑅𝑜 = 0.253 (𝑠𝑒𝑐), (𝐶 = 15 Mbps = 3750 packets/sec) and
(𝑁 = 60 TCP sources) [4]. Using the preceding parameters in (4), the full TCP/AQM system is obtained, and
the transfer function of the model is as (5):
𝑝(𝑠) = (117187.5𝑒−0.253𝑠
) (𝑠2
+ 4.5245𝑠 + 1.9759)
⁄ (5)
3. PROPOSED ENHANCED CONTROLLER
The classical PID controller is considered as a flexible and traditional type of controllers, it’s used for
enhancing system response. Recently, many studies are used these types with a numerous modifications such as
using intelligent techniques such as combines with sliding mode controller or using with any type of neural
network [23], [24], or change its structures to improve the response of the system as in [25], or use the fractional
calculus method by adding an integral or differential fractiona values or together to enhance system behavior [26],
the values are (𝜆 for the integral part and 𝜇 for derivative part), based on this controller gains become five
variables. In this study a nonlinear formula is adopted to improve system performance and led the system to its
efficient response[27], [28] as indicated in Figure 3, it is an enhanced combination between nonlinear gain 𝐾𝑛(𝑒)
and the original PI controller gains, this nonlinear gain is a function of system error also, as shown in (6).
Figure 3. Enhanced PI controller
𝐾𝑛(𝑒) = 𝑇𝑎𝑛ℎ(𝑒) =
𝑒𝑥𝑝(𝑘0𝑒)−𝑒𝑥𝑝(𝑘0𝑒)
𝑒𝑥𝑝(𝑘0𝑒)+𝑒𝑥𝑝(𝑘0𝑒)
(6)
𝑒 = {
𝑒 |𝑒| ≤ 𝑒𝑚𝑎𝑥
𝑒𝑚𝑎𝑥 . 𝑠𝑖𝑔𝑛(𝑒) |𝑒| > 𝑒𝑚𝑎𝑥
(7)
𝐾𝑝, 𝐾𝑖, and 𝐾0 are three parameters that will tuned for achieve an optimal response for this controller.
4. SNAKE OPTIMIZATION ALGORITHM
The SOA was adopted by Hashim and Hussien [29]. It explains the snakes mating way, it describes
how the snakes struggling for finding their suitable associate if there is a good place such as cold weather and
a sufficient food, it’s randomly starting with population generation, and an update for its position is done in
two phases: exploration and exploitation. They start with considring that the male’s numbers and the female’s
numbers are equal when it starts the updating process. This action remains till complete chosen number of the
iteration (𝑇), the first phase is called exploration, it describes the state when there food is not sufficient and the
thier search is done in random style. Quantity of food is calculated using (8) [30].

 ISSN: 1693-6930
1000
𝑄 = 0.5 𝑒𝑥𝑝
𝑡−𝑇
𝑇
(8)
And the equations of this phase are:
𝑋𝑖,𝑚
𝑡+1
= 𝑋𝑟𝑎𝑛𝑑,𝑚
𝑡
± 𝐶2 × 𝐴𝑚 × (𝑋𝑚𝑎𝑥 − 𝑋𝑚𝑖𝑛 ) × 𝑟𝑎𝑛𝑑 + 𝑋𝑚𝑖𝑛 (9)
𝑋𝑖,𝑓
𝑡+1
= 𝑋𝑟𝑎𝑛𝑑,𝑓
𝑡
± 𝐶2 × 𝐴𝑓 × (𝑋𝑚𝑎𝑥 − 𝑋𝑚𝑖𝑛 ) × 𝑟𝑎𝑛𝑑 + 𝑋𝑚𝑖𝑛 (10)
Where 𝑋𝑖,𝑚
𝑡+1
, 𝑋𝑖,𝑓
𝑡+1
represent the 𝑖th places of both males and females. 𝑋𝑟𝑎𝑛𝑑,𝑚
𝑡
, 𝑋𝑟𝑎𝑛𝑑,𝑓,,
𝑡
represent the
places that taken randomly by the two types chosen population (males and females). 𝐶2 represents a known
variable (𝐶2 = 0.5), 𝐴𝑚 is the male ability while 𝐴𝑓 is the female ability to reach to the food place and it is
calculated by (11), (12).
𝐴𝑚 = 𝑒𝑥𝑝
−𝑓𝑟𝑎𝑛𝑑,𝑚
𝑡
𝑓𝑖,𝑚
𝑡+1 (11)
𝐴𝑓 = 𝑒𝑥𝑝
−𝑓𝑟𝑎𝑛𝑑,𝑓
𝑡
𝑓𝑖,𝑓
𝑡+1 (12)
𝐹 explains the fitness function magnitude of males and females. When they find their food, the
exploitation phase is begin, then their places is updated due to their environment temperature (TEMP) and its
founnd as indicated:
𝑇𝐸𝑀𝑃 = exp
−𝑡
𝑇
(13)
When the the weather becomes hot and the temperature exceeds the threshold level, the places of males and
females are updated as shown:
𝑋𝑖,𝑓,𝑚
𝑡+1
= 𝑋𝑓𝑜𝑜𝑑
𝑡
± 𝐶3 × 𝑇𝐸𝑀𝑃 × 𝑟𝑎𝑛𝑑 × (𝑋𝑓𝑜𝑜𝑑 − 𝑋𝑖,𝑓,𝑚
𝑡+1
) (14)
𝑋𝑓𝑜𝑜𝑑
𝑡
is regarded as the optimal place and 𝐶3 equal to 2. Otherwise, the population will replaces between
each other randomly in two modes either fighting or mating and reach the places as explained in (15)-(18):
𝑋𝑖,𝑚
𝑡+1
= 𝑋𝑖,𝑚
𝑡
± 𝐶3 × 𝐹𝑀 × 𝑟𝑎𝑛𝑑 × (𝑋𝑏𝑒𝑠𝑡,𝑓 − 𝑋𝑖,𝑚
𝑡
) (15)
𝑋𝑖,𝑓
𝑡+1
= 𝑋𝑖,𝑓
𝑡+1
± 𝐶3 × 𝐹𝐹 × 𝑟𝑎𝑛𝑑 × (𝑋𝑏𝑒𝑠𝑡,𝑚 − 𝑋𝑖,𝑓
𝑡+1
) (16)
𝑋𝑖,𝑚
𝑡+1
= 𝑋𝑖,𝑚
𝑡
± 𝐶3 × 𝑀𝑚 × 𝑟𝑎𝑛𝑑 × (𝑄 × 𝑋𝑖,𝑓
𝑡
− 𝑋𝑖,𝑚
𝑡
) (17)
𝑋𝑖,𝑓
𝑡+1
= 𝑋𝑖,𝑓
𝑡
± 𝐶3 × 𝑀𝑓 × 𝑟𝑎𝑛𝑑 × (𝑄 × 𝑋𝑖,𝑚
𝑡
− 𝑋𝑖,𝑓
𝑡
) (18)
In (19), (20) is specific for fighting state and [21], [22] is specific for mating state and the 𝐹𝐹, 𝐹𝑀 are
representing the ability of fighting, 𝑀𝑚 and 𝑀𝑓 are considered as a mating ability for males’ and female’s as listed:
𝐹𝑀 = 𝑒𝑥𝑝
−𝑓𝑏𝑒𝑠𝑡,𝑓
𝑓𝑖
(19)
𝐹𝐹 = 𝑒𝑥𝑝
−𝑓𝑏𝑒𝑠𝑡,𝑚
𝑓𝑖
(20)
𝑀𝑚 = 𝑒𝑥𝑝
−𝑓𝑖,𝑓
𝑓𝑖,𝑚
(21)
𝑀𝑓 = 𝑒𝑥𝑝
−𝑓𝑖,𝑚
𝑓𝑖,𝑓
(22)
The selection of worst male and female is happen when egg hatch then switch between them.
𝑋𝑤𝑜𝑟𝑠𝑡,𝑚 = 𝑋𝑚𝑖𝑛 + 𝑟𝑎𝑛𝑑 × (𝑋𝑚𝑎𝑥 − 𝑋𝑚𝑖𝑛 ) (23)
𝑋𝑤𝑜𝑟𝑠𝑡,𝑓 = 𝑋𝑚𝑖𝑛 + 𝑟𝑎𝑛𝑑 × (𝑋𝑚𝑎𝑥 − 𝑋𝑚𝑖𝑛 ) (24)

1001
For monitor system performance in reaching the optimal response one of the cost function relations
will used to monitor the queue size until reachs to its optimal level and achieve stability for the system, the
integral time absolute error (ITAE) [31] was selected for checking system performance, optimal gains of
suggested controller are tuned by SOA tuning method with try to minimizing the magnitude of the fitness
function ITAE used [32], and it is defined in (25) by cheking the system error value continuously.
𝐼𝑇𝐴𝐸 = ∫ 𝑡|e|
∞
0
𝑑𝑡 (25)
Figure 4 indicates the AQM system based on SOA and ,the flowchart of SOA is explained in Figure 5.
Figure 4. AQM system based on SOA
Figure 5. Flowchart of SOA

 ISSN: 1693-6930
1002
5. SIMULATON RESULTS
The simulation results for the AQM system using the suggested controller is presented in this section
using Matlab/Simulink to analyze system performance based on suggested controller and the SOA then
compared it with the two conventional controller (optimal PI and Classical PI controllers). The first controller
is tuned using SOA and the second one is manually tuned, the SOA variables initial values are indicated in
Table 2. The efficient response of the system based on nonlinear proposed controller is appeared clearly when
compared with the two controllers (optimal PI, classical PI) as shown in Figure 6 and the three controllers gains
is shown in Table 3. Then this comparison is analyzed based on the results of response analysis obtained for
the three and it is shown Table 4.
Table 2. SOA parameters
Description Value
Population number 50
Maximum iteration number 30
Variables 3
Figure 6. The AQM system response based on all controllers
Table 3. The three controller’s gains
Controller 𝐾𝑝 𝐾𝐼1 𝐾𝑜
Classical PI 0.0000042 0.000009 -
Optimal PI 0.0000083 0.000074 -
Nonlinear PI 0.0091 0.00451 0.00285
Table 4. Response analysis results for the three controllers
Controller Maximum vershoot (𝑀𝑝 %) Peak time (𝑇𝑝) Rise time (𝑡𝑟) Settling time (𝑡𝑠)
Normal PI 18.33 7.05 2.95 4.4
Optimal PI 7.5 7.15 3.32 4.8
Adaptive PI 0 3.75 1.975 3.4
As indicated in response analysis results in Table 4 the normal PI controller and the optimal PI
controller is similar in there evaluation parameters with small different in their values (355 packets and this led
to overshoot with 18.33% for the normal PI and 322.5 with 7.5% overshoot) while the nonlinear PI controller
is different from these two controllers by its fast response with stable behavior without any overshoot or noise
during simulated time this is due to the nonlinear function and the tuned 𝐾𝑜 variable value within the tanh
hyperbolic function that is used with the PI control which is regulate and stabilize controller behavior, 𝑘𝑜 value
is tuned using SOA algorithm. The fast settling time is appeared on nonlinear PI controller that makes the
system efficiently tracking the desired response, then the analysis of robustness to detect the stability of the
nonlinear controller, two test is done the first one is to monitor its ability in tracking the desired system response
in a stable manner by changing the value of queue size each 50 sec, the propsed controller solves this change
in queue size value regularly and give a stable response in spite of changing desired queue size applied as
shown in Figure 7, while the second test is done by changing the number of TCP sessions from its original
value 𝑁 = 60 to a ± 20% from its original value (+20% equal 72 and -20% equal 48), it can be seen that the
system response is affected when changing system TCP sessions as indicated in Figure 8, when it is increased
to 72 it will suffer from slow response as compared with the original system while when it is decreased to 48
it will suffer from overshoot with value of 8.5 % and needs 10 sec to return to its stable response.

1003
Figure 7. AQM system responses with changing
queue size
Figure 8. AQM system response with changing 𝑁
with ± 20%
6. CONCLUSION
In this paper an intelligent snake optimization tuning algorithm is suggested to tune an enhanced PI
controller with a hyperbolic function to maintain AQM system congestion problem. The controller achieve a
smooth and stable performance for monitoring the system response desired value. A comparison analysis is
utilized with two classical controllers (optimal PI, normal PI) to show the stable and robust response of the
proposed controller based on transient response analysis (peak time, settling time, rise time and overshoot), then
to test system ability to solve problems or changes which may happen during communication a robustness tests
are applied to the system. The results showed an efficient response in saving the desired value of thje queue size
and reach to a stable and robust performance in maintaining the issue of congestion hat occur in AQM system.
REFERENCES
[1] R. Barzamini, M. Shafiee, and A. Dadlani, “Adaptive generalized minimum variance congestion controller for dynamic TCP/AQM
networks,” Computer Communications, vol. 35, no. 2, pp. 170-178, 2012, doi: 10.1016/j.comcom.2011.08.010.
[2] H. A. -Jaber, “An exponential active queue management method based on random early detection,” Journal of Computer Networks
and Communications, 2020, doi: 10.1155/2020/8090468.
[3] B. P. Lee, R. K. Balan, L. Jacob, W. K. G. Seah, and A. L. Ananda, “Avoiding congestion collapse on the Internet using TCP
tunnels,” Computer Networks, vol. 39, no. 2, pp. 207–219, 2002, doi: 10.1016/S1389-1286(01)00311-5.
[4] A. A. Mahawish and H. J. Hassan, “Survey on: A variety of AQM algorithm schemas and intelligent techniques developed for
congestion control,” Indonesian Journal of Electrical Engineering and Computer Science (IJEECS), vol. 23, no. 3, pp. 1419–1431,
2021, doi: 10.11591/ijeecs.v23.i3.pp1419-1431.
[5] M. Z. Al-Faiz and S. A. Sadeq, “Particle Swarm Optimization Based Fuzzy-Neural Like PID Controller for TCP/AQM Router,”
Intelligent Control and Automation, vol. 3, no. 1, 2012, doi: 10.4236/ica.2012.31009.
[6] H. Ashtiani, H. M. Pour, and M. Nikpour, “Active queue management in TCP networks based on fuzzy-PID controller,” Journal of
Applied Computer Science & Mathematics, vol. 6, no. 1, 2012. [Online]. Available: https://jacsm.ro/view/?pid=12_1
[7] M. Z. Al-Faiz and A. M. Mahmood, “Fuzzy-Genetic controller for congestion avoidance in computer networks,” Iraqi Journal of
Computers, Communications, Control, and Systems Engineering, vol. 11, no. 2, pp. 20-27, 2011. [Online]. Available:
https://ijccce.uotechnology.edu.iq/article_46361.html
[8] M. Z. Al-Faiz and S. S. Sabry, “Optimal linear quadratic controller based on genetic algorithm for TCP/AQM router,” 2012
International Conference on Future Communication Networks, 2012, pp. 78-83, doi: 10.1109/ICFCN.2012.6206877.
[9] Q. Xu, G. Ma, K. Ding, and B. Xu, “An Adaptive Active Queue Management Based on Model Predictive Control,” in IEEE Access,
vol. 8, pp. 174489-174494, 2020, doi: 10.1109/ACCESS.2020.3025377.
[10] S. A. Eissa, S. W. Shneen, and E. H. Ali, “Flower Pollination Algorithm to Tune PID Controller of TCP/AQM Wireless Networks”,
Journal of Robotics and Control (JRC), vol. 4, no. 2, 2023, doi: 10.18196/jrc.v4i2.17533.
[11] A. A. A. -Shareha, “Enhanced random early detection using responsive congestion indicators,” International Journal of Advanced
Computer Science and Applications (IJACSA), vol. 10, no. 3, 2019, doi: 10.14569/IJACSA.2019.0100347.
[12] M. M. Abualhaj, A. A. A. -Shareha, and M. M. Al-Tahrawi, “FLRED: an efficient fuzzy logic based network congestion control
method,” Neural Computing and Applications, vol. 30, pp. 925-935, 2018, doi: 10.1007/s00521-016-2730-9.
[13] A. Alsaaidah, M. Zalisham, M. Fadzli, and H. A. -Jaber, “Markov-modulated bernoulli-based performance analysis for gentle BLUE
and BLUE algorithms under bursty and correlated traffic,” Journal of Computer Science, vol. 12, no. 6, pp. 289-299, 2016,
doi: 10.3844/jcssp.2016.289.299.
[14] C. Brandauer, G. Iannaccone, C. Diot, T. Ziegler, S. Fdida, and M. May, “Comparison of tail drop and active queue management
performance for bulk-data and Web-like Internet traffic,” Proceedings. Sixth IEEE Symposium on Computers and Communications,
2001, pp. 122-129, doi: 10.1109/ISCC.2001.935364.
[15] M. Welzl, Network Congestion Control, Managing Internet Traffic, England : John Wiley & Sons, Ltd, 2005. [Online]. Available:
http://pws.npru.ac.th/sartthong/data/files/Traffic__Wiley_Series_on_Communications_Networking.pdf
[16] M. K. Oudah, M. Q. Sulttan, and S. W. Shneen, “Fuzzy type 1 PID controllers design for TCP/AQM wireless networks,” Indonesian
Journal of Electrical Engineering and Computer Science, vol. 21, no. 1, pp. 118-127, 2021, doi: 10.11591/ijeecs.v21.i1.pp118-127.
[17] 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,” SIGCOMM '00: Proceedings of the conference on Applications, Technologies, Architectures, and Protocols
for Computer Communication, 2000, pp. 151-160, doi: 10.1145/347059.347421.
[18] S. K. Bisoy, P. K. Pattnaik, M. Sain, and D. -U. Jeong, “A Self-Tuning Congestion Tracking Control for TCP/AQM Network for
Single and Multiple Bottleneck Topology,” IEEE Access, vol. 9, pp. 27723-27735, 2021, doi: 10.1109/ACCESS.2021.3056885.

 ISSN: 1693-6930
1004
[19] S. S. Sabry and N. M. Kaittan, “Grey wolf optimizer based fuzzy-PI active queue management design for network congestion
avoidance,” Indonesian Journal of Electrical Engineering and Computer Science (IJEECS), vol. 18, no. 1, pp. 199-208, 2020,
doi: 10.11591/ijeecs.v18.i1.pp199-208.
[20] C. V. Hollot, V. Misra, D. Towsley, and W. -B. Gong, “A control theoretic analysis of RED,” Proceedings IEEE INFOCOM 2001.
Conference on Computer Communications. Twentieth Annual Joint Conference of the IEEE Computer and Communications
Society, 2001, vol. 3, pp. 1510-1519, doi: 10.1109/INFCOM.2001.916647.
[21] L. H. Abood, B. K. Oleiwi, A. J. Humaidi, A. A. Al-Qassar, and A. S. M. Al-Obaidi, “Design a robust controller for congestion
avoidance in TCP/AQM system,” Advances in Engineering Software, vol. 176, 2023, doi: 10.1016/j.advengsoft.2022.103395.
[22] M. I. Berbek and A. A. Oglah, “Adaptive neuro-fuzzy controller trained by genetic-particle swarm for active queue management in
internet congestion,” Indonesian Journal of Electrical Engineering and Computer Science (IJEECS), vol. 26, no. 1, pp. 229-242,
2022, doi: 10.11591/ijeecs.v26.i1.pp229-242.
[23] L. Hattim, E. H. Karam, and A. H. Issa, “Implementation of Self Tune Single Neuron PID Controller for Depth of Anesthesia by
FPGA,” in New Trends in Information and Communications Technology Applications: Third International Conference, NTICT
2018, 2018, pp. 159-170, doi: 10.1007/978-3-030-01653-1_10.
[24] W. E. A. -Lateef, Y. N. I. Alothman, and S. A. -H. Gitaffa, “An optimal motion path planning control of a robotic manipulator based
on the hybrid PI-sliding mode controller,” Bulletin of Electrical Engineering and Informatics (BEEI), vol. 12, no. 2, pp. 727-737,
2023, doi: 10.11591/eei.v12i2.3968.
[25] H. I. Ali, and A. H. Saeed, “Robust PI-PD Controller Design for Systems with Parametric Uncertainties,” Engineering and
Technology Journal, vol. 34, no. 11, pp. 2167–2173, 2016, doi: 10.30684/etj.34.11A.20.
[26] L. H. Abood and R. Haitham, “Design an Optimal Fractional Order PI Controller for Congestion Avoidance in Internet Routers,”
Mathematical Modelling of Engineering Problems, vol. 9, no. 5, pp. 1321–1326, 2022, doi: 10.18280/mmep.090521.
[27] D. Pathak, S. Bhati, and P. Gaur, “Fractional‐order nonlinear PID controller based maximum power extraction method for a direct‐driven
wind energy system,” International Transactions on Electrical Energy Systems, vol. 30, no. 12, 2020, doi: 10.1002/2050-7038.12641.
[28] E. H. Ali, A. H. Reja, and L. H. Abood, “Design hybrid filter technique for mixed noise reduction from synthetic aperture radar imagery,”
Bulletin of Electrical Engineering and Informatics (BEEI), vol. 11, no. 3, pp. 1325-1331, 2022, doi: 10.11591/eei.v11i3.3708.
[29] F. A. Hashim and A. G. Hussien, “Snake Optimizer: A novel meta-heuristic optimization algorithm,” Knowledge-Based Systems,
vol. 242, 2022, doi: 10.1016/j.knosys.2022.108320.
[30] M. Rawa, “Towards Avoiding Cascading Failures in Transmission Expansion Planning of Modern Active Power Systems Using
Hybrid Snake-Sine Cosine Optimization Algorithm,” Mathematics, vol. 10, no. 8, 2022, doi: 10.3390/math10081323.
[31] L. H. Abood, “Optimal modified PID controller for automatic voltage regulation system,” in AIP Conference Proceedings, 2022,
vol. 2415, doi: 10.1063/5.0092583.
[32] L. H. Abood, N. N. Kadhim, and Y. A. Mohammed, “Dual stage cascade controller for temperature control in greenhouse,” Bulletin
of Electrical Engineering and Informatics (BEEI), vol. 12, no. 1, pp, 51-58, 2023, doi: 10.11591/eei.v12i1.4328.
BIOGRAPHIES OF AUTHORS
Yousra Abd Mohammed is a lecturer at the Communication Engineering
Department, University of Technology-Baghdad, Iraq since 2005. She received her B.Sc. in
Electronic and Communication Engineering from Technology University/Baghdad, Iraq in
1992 and her M.Sc. degree in Computer Engineering from Technology University/Baghdad
in 2004. Her research interests include Control Systems, Encryption and Decryption
Algorithms. She can be contacted at email: yousra.a.mohammed@uotechnology.edu.iq.
Layla H. Abood received her B.Eng. and M.Sc. and PHD degrees in Electronic and
Communication Engineering from the University of Technology- Baghdad. She is currently an
Academic staff member in the Department of Control and System Engineering, University of
Technology, Baghdad, Iraq. Her research interests are Artificial Intelligent, Control Systems,
System Modeling, Optimization Techniques, IoT, Image processing, Robotics, Microcontrollers,
FPGA and Embedded systems. She can be contacted at email: 60066@uotechnology.edu.iq.
Nahida Naji Kadhim received B.Sc. and M.Sc. degrees in Control Engineering
and Mechatronics Engineering from Control and Systems Engineering Department in 1993
and 2005 respectively, University of Technology-Baghdad, Iraq. She is currently an academic
staff member in the Department of Control and Systems engineering. University of
Technology, Baghdad, Iraq. Her researcher interests are Control System, Robotic System,
Identification System and Artificial Inelegant. She can be contacted at email:
nahida.n.kadhim@uotechnology.edu.iq.

Design and simulation an optimal enhanced PI controller for congestion avoidance in TCP/AQM system

More Related Content

Similar to Design and simulation an optimal enhanced PI controller for congestion avoidance in TCP/AQM system (20)

More from TELKOMNIKA JOURNAL (20)

Recently uploaded (20)

Design and simulation an optimal enhanced PI controller for congestion avoidance in TCP/AQM system