1 / 20

Algoritmi de combatere a congestiei

Algoritmi de combatere a congestiei. Profesor : Stefan Stancescu Student: Toma Oana Madalina. Congestie-Introducere.

nate
Download Presentation

Algoritmi de combatere a congestiei

An Image/Link below is provided (as is) to download presentation Download Policy: Content on the Website is provided to you AS IS for your information and personal use and may not be sold / licensed / shared on other websites without getting consent from its author. Content is provided to you AS IS for your information and personal use only. Download presentation by click this link. While downloading, if for some reason you are not able to download a presentation, the publisher may have deleted the file from their server. During download, if you can't get a presentation, the file might be deleted by the publisher.

E N D

Presentation Transcript

  1. Algoritmi de combatere a congestiei Profesor: Stefan Stancescu Student: TomaOanaMadalina

  2. Congestie-Introducere • Congestia este procesul de pierdere a pachetelor, transmise prin retea, din cauza lipsei de spatiu in buffere-le de stocare a concentratorilor instalati in reteaua respectiva. Presupun ca avem un segment de retea ca in figura de mai jos:

  3. Algoritmi de combatere a congestiei • Conform [24] algoritmii de combatere a congestiei se clasifica in: Figure 2. Controlulcongestiei in buclainchisa. Algoritmul RED • Conform acesteiclasificarialgoritmii AQM fac parte din algoritmii closed-loop.

  4. Algoritmul RED

  5. Algoritmul RED (2) Figure 3. Algoritmul Red aplicat buffer-ului unui singur router [2]

  6. Algoritmul GRED

  7. ARED (Adaptive RED)

  8. NLRED (Nonlinear RED) Figure 6. Functia de pierdere a pachetelor pentru NLRED [3]

  9. DRED (Dynamic RED) Figure 7. Functia de pierdere a pachetelor pentru DRED [3]

  10. IRED (Improved RED)

  11. RDRED (Real-time Dynamic RED) • Celemaimultevariante ale algoritmului RED incamaiutilizeazalungimeamedie a coziicamasura a congestiei, daracestlucru nu poatereflecta rapid evolutialungimiicozii. • Lungimeainstantanee a cozii se folosestecamasura a congestiei, in loc de lungimeamedei a cozii, siacestmijloc de control al algoritmului RED modificat a fostdenumit Real-Time Dynamic RED (RDRED).

  12. RDRED (Real-time Dynamic RED)

  13. Compararea algoritmilor • Simularea s-a realizat cu o retea simpla bottleneck ce contine doua routere si N perechi de sender si noduri sink. Toate link-urile au fost configurate la 10Mbps si 10ms. Au fost implementati urmatorii algoritmi AQM pentru routere: RED, ARED, NLRED, DRED, IRED si RDRED. In tabelul 1 sunt date valorile parametrilor pentru toti algoritmii testati:

  14. Compararea algoritmilor (2)

  15. Compararea algoritmilor (4)

  16. Compararea algoritmilor (4)

  17. Performantele in conditii de traficdinamic • Traficul TCP estesetat la 100 la inceput. S-au adaugatdouagrupuri de trafic, fiecaredintreeleconsista in 100 de flow-uri TCP, elesunt activate la 100s si 200s ramanand active pana la 300s, respectiv 400s. • Evolutialungimiicoziiesteilustrata in figura 10. • Lungimeacozii de asteptaredupaaplicareaalgoritmilor RED, ARED si NLRED variaza cu schimbareaincarcariitraficului. • Lungimilecozilor de asteptareoscileazafoartemult, producandastfelintarzieri ale pachetelor. • DRED, IRED si RDRED se comporta bine, chiarsi la variatiibruste ale traficului. • Lungimeacozii de asteptare a algoritmului RDRED atingeceamaibunaperformanta.

  18. Figure 10. Evolutia lungimii cozii in conditii de trafic dinamic [3]

  19. Concluzii • Din rezultatelesimularii, se poateobservaca RED, ARED, NLRED nu se comportafoarte bine in conditii de incarcarevariabila a retelei. • Este greu de stabilizatlungimeacozii de asteptarepentru a avea o bunaperformanta la o gamavariata de nivele de incarcare. • DRED, IRED si RDRED pot stabilizafoarte bine lungimeacozii, de aicireiese o intarziereacceptabila. • RDRED atingeceamaibunaperformantadintretotialgoritmiitestati, petipurile de retelealese in simulare. • IRED si RDRED suntsimplu de implementatsiestefoarteeficientapentru a introduce lungimeainstantanee a cozii de asteptare, camasura a congestiei.

  20. Bibliografie • [1] http://rast.orgfree.com/retele/retele_2002_07.html • [2] OmidSeifaddini , Azizol Abdullah and Hamid Vosough “RED, GRED, AGRED CONGESTION CONTROL ALGORITHMS IN HETEROGENEOUS TRAFFIC TYPES “4th International Conference on Computing and Informatics, ICOCI 2013 28-30 August, 2013 Sarawak, Malaysia. Universiti Utara Malaysia • [3] Minjuan Cheng, Xiaoming Ma, Performance Evaluation of Queue Management Methods for Congestion Control, Journal of Information & Computational Science 9: 6 (2012) 1599–1608 • [4] K. Zhou, K. L. Yeung, V. O. K. Li, Nonlinear RED: A simple yet efficient active queue management • scheme, Computer Networks, 50(18), 2006, 3784-3794 • [5] M. Cheng, H. Wang, L. Yan, Dynamic RED: A modified random early detection, Journal of • Computational Information Systems, 7(14), 2011, 5243-5250  • [6] H. Wang, Z. Ye, B. Wang, Using auto-tuning proportional integral probability to improve random • early detection, Proc. the 13th IEEE International Conference on Communication Technology • (ICCT), 2011, 1107-1111 • [24] A. Tanenbaum, Computer networks 5th edition, Practice Hall, USA, 2011

More Related