NETWORK LAYER INTERNET TOPOLOGY CONSTRUCTION M. Engin Tozal & - - PowerPoint PPT Presentation

network layer internet topology construction
SMART_READER_LITE
LIVE PREVIEW

NETWORK LAYER INTERNET TOPOLOGY CONSTRUCTION M. Engin Tozal & - - PowerPoint PPT Presentation

1 NETWORK LAYER INTERNET TOPOLOGY CONSTRUCTION M. Engin Tozal & Kamil Sarac The University of Texas at Dallas AGENDA 2 Internet Topology and its Representation Existing Topology Representations Subnet Level Topologies


slide-1
SLIDE 1

NETWORK LAYER INTERNET TOPOLOGY CONSTRUCTION

  • M. Engin Tozal & Kamil Sarac

The University of Texas at Dallas

1

slide-2
SLIDE 2

AGENDA

 Internet Topology and its Representation  Existing Topology Representations  Subnet Level Topologies  Subnet Inference with XNET  Current Research

2

slide-3
SLIDE 3

Internet Topology

 Internet Topology Maps

 Represented as a graph G(V,E) such that V is the set of

  • bjects and E is the set of links between objects in V.

 Vertices:

 Autonomous Systems (ASes)  Routers  Router Triangles  Interfaces  Subnetworks (Subnets)

3

slide-4
SLIDE 4

Internet Topology

 Inter-Connections

 Policy-based connections  Subnets  Routers

 Do we really have a distinction between components in

the Internet and their inter-connections?

 Or is it a matter of representation and interest?

4

slide-5
SLIDE 5

AGENDA

 Internet Topology and its Representation  Existing Topology Representations  Subnet Level Topologies  Subnet Inference with XNET  Current Research

5

slide-6
SLIDE 6

Existing Internet Topologies

 AS Level  Interface Level  Router Level

6

slide-7
SLIDE 7

Existing Internet Topologies AS Level

AS1 AS2 AS3 AS4 A Sample Section of the Internet Topology at the Network Layer

7

slide-8
SLIDE 8

Existing Internet Topologies AS Level

AS2 AS1 AS3 AS4 AS Level Representation

8

slide-9
SLIDE 9

Existing Internet Topologies Interface Level

AS1 AS2 AS3 AS4

X Z Y

A Sample Section of the Internet Topology at the Network Layer

9

slide-10
SLIDE 10

Existing Internet Topologies Interface Level

Interface Level Representation

X Z Y

10

slide-11
SLIDE 11

Existing Internet Topologies Router Level

A Sample Section of the Internet Topology at the Network Layer

11

R1 R3 R2 S5 S4 S3 S2 S1 R4 R6 R5 R7 R8

slide-12
SLIDE 12

Existing Internet Topologies Router Level

Router Level Representation

12

R1 R3 R2 R4 R6 R5 R7 R8

slide-13
SLIDE 13

AGENDA

 Internet Topology and its Representation  Existing Topology Representations  Subnet Level Topologies  Subnet Inference with XNET  Current Research

13

slide-14
SLIDE 14

Subnet Level Topologies

A Sample Section of the Internet Topology at the Network Layer

14

R1 R3 R2 S5 S4 S3 S2 S1 R4 R6 R5 R7 R8

slide-15
SLIDE 15

Subnet Level Topologies

15

S5 S4 S3 S2 S1

slide-16
SLIDE 16

Subnet Level Topologies

Subnet Level Representation

16

S5 S4 S3 S2 S1 /30 /29 /31 /29 /31

slide-17
SLIDE 17

Subnet Level Topologies

 Subnet Level Internet Topologies

 A subnet S is defined by the set of interfaces that it

accommodates

 A vertex in the graph is a subnet  A link between two subnets represents the router that

directly connects two subnets to each other

 TraceNET is a tool for building subnet level Internet

topologies

 XNET is another tool for inferring individual subnets

17

slide-18
SLIDE 18

Utility of Subnet Level Topologies

18

 Building node-and-link disjoint end-to-end paths for

  • verlay network design
slide-19
SLIDE 19

Utility of Subnet Level Topologies

 Studying Subnet Level Topology Characteristics

 Degree Distribution  IP address space Utilization  Betweenness  Assortativity  Clustering Coefficient  and so on…

19

slide-20
SLIDE 20

AGENDA

 Internet Topology and its Representation  Existing Topology Representations  Subnet Level Topologies  Subnet Inference with XNET  Current Research

20

slide-21
SLIDE 21

Subnet Inference with XNET

 ExploreNET (XNET)

 Given an IP address t, XNET builds the subnet S hosting

t

 Collects all alive IP addresses accommodated by S  Labels S by its observed subnet mask  TraceNET

 Given a destination address d

 Returns a list of subnets appearing between the source and

the destination address d

21

slide-22
SLIDE 22

Subnet Inference with XNET

 XNET vs TraceNET

 Both are based on the same principles  TraceNET has more data to draw inference (subnets

and IP addresses appearing in previous hops)

 XNET can be utilized in uniform random subnet

sampling

 TraceNET sampling possesses source dependency bias

22

slide-23
SLIDE 23

Subnet Inference with XNET

23

A S

slide-24
SLIDE 24

Subnet Inference with XNET

 Scope Delimitation Test  Far Fringe Interface Detection Test  Ingress Fringe Interface Detection Test  Close Fringe Interface Detection Test

24

slide-25
SLIDE 25

Subnet Inference with XNET

Fringe Interface Detection

25

A

Far-fringe interfaces Close-fringe interfaces Ingress-fringe interfaces

S

slide-26
SLIDE 26

AGENDA

 Internet Topology and its Representation  Existing Topology Representations  Subnet Level Topologies  Subnet Inference with XNET  Current Research

26

slide-27
SLIDE 27

Current Research

 Network Level Internet Topology Maps  A mathematical framework for sampling subnets

using XNET

 Developing a network generation model based on

subnets

27

slide-28
SLIDE 28

Internet Topology Representations

Example Internet Topology

28

slide-29
SLIDE 29

Internet Topology Representations

Router Level Representation

29

slide-30
SLIDE 30

Internet Topology Representations

Subnet Level Representation

30

slide-31
SLIDE 31

Internet Topology Representations

Network Level Representation

31

slide-32
SLIDE 32

Discussion

 Questions & Comments

32

slide-33
SLIDE 33

Evaluations

 Characteristics of subnets in six geographically

disperse Tier-2 ISP networks

 PCCW Global (ISP-1)  nLayer (ISP-2)  France Telecom (ISP-3)  Telecom Italia Sparkle (ISP-4)  Interroute (ISP-5)  MZIMA (ISP-6)

33

slide-34
SLIDE 34

Evaluations

34

slide-35
SLIDE 35

Evaluations

35

slide-36
SLIDE 36

Evaluations

 Power Law in Prefix Length Distribution

36

slide-37
SLIDE 37

Evaluations

 Power Law in Prefix Length Distribution

37

slide-38
SLIDE 38

Subnet Inference with XNET

38