Congestion Control in Recursive Network Architectures David Hayes, - - PowerPoint PPT Presentation

congestion control in recursive network architectures
SMART_READER_LITE
LIVE PREVIEW

Congestion Control in Recursive Network Architectures David Hayes, - - PowerPoint PPT Presentation

Congestion Control in Recursive Network Architectures David Hayes, Peyman Teymoori, and Michael Welzl University of Oslo [davihay , peymant , michawe]@ifi . uio . no IETF95 Buenos Aires ICCRG David IETF95 Buenos Aires ICCRG 1 / 9


slide-1
SLIDE 1

Congestion Control in Recursive Network Architectures

David Hayes, Peyman Teymoori, and Michael Welzl University of Oslo [davihay,peymant,michawe]@ifi.uio.no IETF95 Buenos Aires — ICCRG

David IETF95 Buenos Aires — ICCRG 1 / 9

slide-2
SLIDE 2

Pristine: Who we are and what we’re doing

ict-pristine.eu

RINA

Recursive Inter-Network Architecture tools RINA simulator

Omnet++

RINA sdk

C++

Research Security Congestion Control Routing Scheduling and QoS Standards Pouzin Society ISO

David IETF95 Buenos Aires — ICCRG 2 / 9

slide-3
SLIDE 3

The Internet and Recursive Architectures

Internet Application TCP/UDP IP Link LLC MAC Physical QoS, MPLS, etc Proxies, middle boxes, etc Recursive Architecture Application Generic Layer ... Generic Layer Physical

Examples include: Recursive Inter-Network Architecture (RINA) http://www.pouzinsociety.org/ Recursive Network Architecture (RNA) http://www.isi.edu/rna/ David IETF95 Buenos Aires — ICCRG 3 / 9

slide-4
SLIDE 4

Comparative RINA Example

S1 R1

Router1

TCP

L1 L2 L3 L4 L1 L2 L3 L1 L2 L3 L4

Split TCP

L1 L2 L3 L4 L1 L2 L3 L4 L1 L2 L3 L4

RINA

Stack 2-DIF 1-DIF 1-DIF

EFCP RMT EFCP RMT EFCP RMT EFCP RMT EFCP RMT EFCP RMT EFCP RMT L1 L2 L1 L2 L1 L2 From P. Teymoori, M. Welzl, G. Stein, E. Grasa, R. Riggio, K. Rausch, and D. Siracuss. Congestion control in the recursive internetwork architecture (RINA). In IEEE International Conference on Communications (ICC), Next Generation Networking and Internet Symposium, May 2016. to appear David IETF95 Buenos Aires — ICCRG 4 / 9

slide-5
SLIDE 5

Congestion Control

Types of “solvable” Network Congestion Network adaption – traffic rerouted along less congested paths Network restriction – admission control and policing End system adaption – traffic sources adjust send rates Recursive CC work in RINA Stability of chained and stacked congestion controllers Congestion signals and how they influence performance Stable scalable new types of congestion control

David IETF95 Buenos Aires — ICCRG 5 / 9

slide-6
SLIDE 6

Recursive Congestion Signals

ISP 1 Access ISP 2 Access GW GW Public Internet Layer Inter ISP Layer ISP 1 Layer ISP 2 Layer Server Host

David IETF95 Buenos Aires — ICCRG 6 / 9

slide-7
SLIDE 7

Recursive Congestion Signals

ISP 1 Access ISP 2 Access GW GW Public Internet Layer Inter ISP Layer ISP 2 Layer Server Host ISP 1 Layer

X

David IETF95 Buenos Aires — ICCRG 6 / 9

slide-8
SLIDE 8

Recursive Congestion Signals

ISP 1 Access ISP 2 Access GW GW Public Internet Layer Inter ISP Layer ISP 1 Layer ISP 2 Layer Server Host

X

Feedback options in RINA Implicit:

◮ Push back and push up

Explicit:

◮ Packet marking and

reflection

◮ Control packet notifications David IETF95 Buenos Aires — ICCRG 6 / 9

slide-9
SLIDE 9

Recursive Congestion Signals

ISP 1 Access ISP 2 Access GW GW Public Internet Layer Inter ISP Layer ISP 1 Layer ISP 2 Layer Server Host

X

Feedback options in RINA Implicit:

◮ Push back and push up

Explicit:

◮ Packet marking and

reflection

◮ Control packet notifications

Potentially provides Faster responses Smaller buffer requirements Cautions — Current research CC interactions Stability

David IETF95 Buenos Aires — ICCRG 6 / 9

slide-10
SLIDE 10

Stable CC inspired by Logistic Growth in Nature

2 4 6 8 K t Population, P(t) P(t) =

KP0ert K+P0(ert−1)

P.-F. Verhulst. Notice sur la loi que la population poursuit dans son accroissement. Correspondance math´ ematique et physique, 10:113–121, 1838. URL http://books.google.com/?id=8GsEAAAAYAAJ

David IETF95 Buenos Aires — ICCRG 7 / 9

slide-11
SLIDE 11

Logistic Growth Based Congestion Control

Competition for resources incrementally for senders: xi(t + 1) = xi(t)ri  C − xi(t) − N − 1 N

N

  • j=1

xj   + xi(t) Recursive layered congestion control – ongoing predator–prey food chain based CC has calculable convergence and stability properties

David IETF95 Buenos Aires — ICCRG 8 / 9

slide-12
SLIDE 12

Conclusions and Future work

Stability and CC interactions also relate to the current Internet New CC can work in the Internet as well Upcoming publication:

  • P. Teymoori, M. Welzl, G. Stein, E. Grasa, R. Riggio, K. Rausch, and D. Siracuss. Congestion control in the

recursive internetwork architecture (RINA). In IEEE International Conference on Communications (ICC), Next Generation Networking and Internet Symposium, May 2016. to appear ◮ RINA congestion control in data centers ◮ Congestion control signals and stability

OCARINA (Optimizations to Compel Adoption of RINA) (see https://titan.uio.no/node/1403)

Acknowledgements The authors are funded by the European Community under its Seventh Framework Programme through the PRISTINE project (CNECT-ICT-619305). The views expressed are solely those of the authors. For more information about PRISTINE see http://ict-pristine.eu/

David IETF95 Buenos Aires — ICCRG 9 / 9

slide-13
SLIDE 13

Extra slides

David IETF95 Buenos Aires — ICCRG 1 / 1