Optimizing PID TCP/AQM using nelder-mead simplex approach

Fajri, Misbahul; Ramli, Kalamullah

doi:10.1145/3162957.3163017

Cited by 4 publications

(5 citation statements)

References 10 publications

Supporting

Mentioning

Contrasting

Unclassified

Order By: Relevance

“…Random Early Detection (RED) [7] algoritma adalah AQM yang cukup terkenal dan banyak digunakan pada router, yang dikembangkan untuk mengatasi kekurangan tradisional AQM yaitu metode FIFO/Drop Tail. Skema lain yang dikembangkan oleh peneliti menggunakan pengendali PID dengan optimasi parameternya [8], yang dikemudian diimplemetasikan ke NS2 dengan pengendali PI yang dioptimasi [9].…”

Section: Pendahuluanunclassified

Implementasi Active Queue Management pada Jaringan Bottleneck

Fajri

Ratnasari²,

Prabowo³

2023

InComTech

View full text Add to dashboard Cite

Teknologi TCP end-to-end congestion control hanya menangani kemacetan jaringan dari suatu host tetapi tidak pada jaringan atau intermediate network, oleh karena itu teknologi Active Queue Management (AQM) dikembangkan untuk mengantisipasi kemacetan didalam jaringan yaitu pada router. AQM dapat mengatasi buffer overflow pada router dan mencegah terjadinya masalah sinkronisasi pada jaringan yang padat, yaitu dengan mendeteksi sebelum terjadi kemacetan jaringan. AQM RED (random early detection) dikembangkan dengan pendekatan heuristik untuk mengantikan kekurangan mekanisme Droptail pada buffer router, sehingga dapat mengatisipasi memburuknya kondisi jaringan yaitu; hilangnya paket yang berlebihan, utilitasasi link yang rendah (low throughput), dan tingginya antrian (delay), dengan demikian kemacetan yang parah (congestion collapse) tidak akan terjadi. Permasalahan ini menarik untuk dipelajari dampak implementasi AQM yang akan dilakukan dalam penelitian ini. Protokol AQM adalah mekanisme pengendali kemacetan yang diterapkan pada lapisan ke-tiga OSI pada peralatan jaringan router, dengan menggunakan simulator NS2 dibuat topologi jaringan bottleneck dan diuji coba dengan skenario trafik. Pada penelitian ini akan diimplementasikan metode AQM RED yang disimulasikan dengan pengiriman paket FTP dengan beberapa skenario trafik. Hasilnya berupa karakteristik jaringan dapat dilihat yaitu; jumlah paket yang ada pada buffer router, lamanya antrian paket pada buffer , dan banyakanya paket yang dibuang, sehingga dapat dianalisa kinerja RED yang merupakan tujuan penelitian ini. Kesimpulan yang didapat menunjukan bahwa RED mempunyai keunggulan dari karakteristik delay yang rendah dibandingkan mekanisme tanpa AQM atau Drop-tail.

show abstract

Section: Pendahuluanunclassified

Implementasi Active Queue Management pada Jaringan Bottleneck

Fajri

Ratnasari²,

Prabowo³

2023

InComTech

View full text Add to dashboard Cite

show abstract

“…Through the control theory approach, AQM can be designed and analyzed practically such as PI [4], 2DoF PID [13], PI-PD [14] and PID AQM [15] solution to overcome lacks of heuristic approach. Likewise, another work [12] uses an optimal control method to tune parameters of PID AQM. It showed that the structure type of PID can be deployed in communication networks.…”

Section: Related Workmentioning

confidence: 99%

“…In [10,11] regulate queue length with adaptive PI with neuron based, leads to high computation and hard in implementation along with heuristic optimization algorithm. Other research [12] was to obtain PID parameters using an optimization approach and to suggest to develop with different structure of PID controllers to implement it with an adaptive mechanism because many AQMs have problems with robustness, especially in changing traffic conditions and effectively handling bursty traffic. To overcome this, we propose an adaptive queue ratio proportional integral (ARPI) regulates queue length to the reference value.…”

Section: Introductionmentioning

confidence: 99%

Adaptive Active Queue Management based on Queue Ratio of Set-point Weighting

Fajri

Ramli

2022

Int. j. commun. netw. inf. secur.

View full text Add to dashboard Cite

Presently, active queue management (AQM) is one of the important considerations in communication networks. The challenge is to make it simple and robust in bursty traffic and uncertain network conditions. This paper proposes a new AQM scheme, an adaptive ratio proportional integral (ARPI), for adaptively controlling network congestion in dynamic network traffic conditions. First, AQM was designed by adding a set-point weighting structure to a proportional integral (PI) controller to reduce the burstiness of network traffic. Second, an adaptive set-point weighting based on the ratio of instantaneous queue length to the set-point queue and the buffer size was proposed to improve the robustness of a non-linear network. The proposed design integrates the aforementioned expectations into one function and needs only one parameter change to adapt to fluctuating network condition. Hence, this scheme provides lightweight computation and simple software and hardware implementation. This approach was analyzed and compared with the PI AQM scheme. Evaluation results demonstrated that our proposed AQM can regulate queue length with a fast response, good stability under any traffic conditions, and small queuing delay.

show abstract

“…Sabry and Kaittan [30] suggest used grey wolf optimizer (GWO) algorithm with a fuzzy-PI controller as an active queue management (AQM) for controls congestion in internet routers and for fulfilling faster response and reduce the steady-state error, the simulation results showed good stability for the proposed controller with the ability to reach a fast response with different loads for the network. Fajri and Ramli [31] are proposed to use a nelder-mead simplex (NMS) method to get the best parameters of the PID controller to enhance the work of TCP/AQM router system, the results of this work appeared the effectiveness of combination between NMS and integral of time-weighted absolute error (ITAE) criterion on other criteria in sense of control for congestion in a network. Li et al [32] are considered a novel AQM scheme for TCP network, when they using integral back-stepping technique (IB) and minimax method with a suitable condition and conformable control, the simulation results showed are not only used to deal with perturbations produced by user datagram protocol (UDP) streams but the convergent time of the signals can also be shortened.…”

Section: Introductionmentioning

confidence: 99%

BBO tuned PI controller for the stability of TCP networks

et al. 2023

View full text Add to dashboard Cite

The congestion is the most important issue that effects on the performance of data transition over internet networks. One of the important techniques developed is active queue management (AQM) that prepares an efficient congestion control by reducing losing packets, queue size, and energy consumption. Therefore, AQM technique deemed as a base of many congestion control algorithms schemes. This work suggested the using of proportional integral (PI) controller as an AQM and then use an optimized control system such as biogeography-based optimization (BBO) with PI controller as (BBO-PI). The optimal control (BBO-PI) is characterized by access to design and fine-tuning of defining the shapes of the optimal parameters of PI controller. The BBO algorithm was implemented by using the mathematical system model by M-file/Matlab and Simulink. The simulation results showed the best performance for the transmission control protocol (TCP) network when compared the system with using the PI controller and using optimal control (BBO-PI), the ratio of enhancing the system with using of BBO-PI better than using a PI controller only in terms of rising time is 1.11, settling time is 2.85 and overshooting is 85%. Therefore, the proposed method was very fast and required few iterations.

show abstract

Optimizing PID TCP/AQM using nelder-mead simplex approach

Cited by 4 publications

References 10 publications

Implementasi Active Queue Management pada Jaringan Bottleneck

Implementasi Active Queue Management pada Jaringan Bottleneck

Adaptive Active Queue Management based on Queue Ratio of Set-point Weighting

BBO tuned PI controller for the stability of TCP networks

Contact Info

Product

Resources

About