Multiple Source, Multiple Destination Network Tomography Michael Rabbat IEEE Infocom, Hong Kong...

24
Multiple Source, Multiple Destination Network Tomography Michael Rabbat IEEE Infocom, Hong Kong Wednesday, March 10, 2004 Co-Authors: Mark Coates and Robert Nowak

Transcript of Multiple Source, Multiple Destination Network Tomography Michael Rabbat IEEE Infocom, Hong Kong...

Page 1: Multiple Source, Multiple Destination Network Tomography Michael Rabbat IEEE Infocom, Hong Kong Wednesday, March 10, 2004 Co-Authors: Mark Coates and Robert.

Multiple Source, Multiple Destination Network Tomography

Michael Rabbat

IEEE Infocom, Hong Kong

Wednesday, March 10, 2004

Co-Authors:

Mark Coates and Robert Nowak

Page 2: Multiple Source, Multiple Destination Network Tomography Michael Rabbat IEEE Infocom, Hong Kong Wednesday, March 10, 2004 Co-Authors: Mark Coates and Robert.

What is Network Tomography?

Logical Topology

A

1 2 3

Goal: Characterize the internal network using end-to-end measurements

6 754

32

1

Good Bad Ugly

+Link-level Performance Parameters

Page 3: Multiple Source, Multiple Destination Network Tomography Michael Rabbat IEEE Infocom, Hong Kong Wednesday, March 10, 2004 Co-Authors: Mark Coates and Robert.

Back-to-Back Packet Probes

A

1 2

Similar experience

Independent experiences

(Keshav, ’91) (Carter & Crovella, ’96)

Repeat and average) Take T measurements

Independent behavior onunshared links allows us to separate performance effects(e.g., loss, delay) on the different branches

Page 4: Multiple Source, Multiple Destination Network Tomography Michael Rabbat IEEE Infocom, Hong Kong Wednesday, March 10, 2004 Co-Authors: Mark Coates and Robert.

Reconstruct The Network (Single Source)

1.5

0.5

1.0 2.5

1.0

2.0

1.0 2.5 1.5

1.5

1.0 3.0 1.0

1.5

• Link-level characteristics (loss, delay) estimation• Network topology identification

Tightly coupled problems

(Duffield, Towsley et al., ’99) (Coates & Nowak, ’00) (Byers et al., ’00)

Page 5: Multiple Source, Multiple Destination Network Tomography Michael Rabbat IEEE Infocom, Hong Kong Wednesday, March 10, 2004 Co-Authors: Mark Coates and Robert.

Probe From Multiple HostsA

1 2 3

B(Bu et al., ’02)

Page 6: Multiple Source, Multiple Destination Network Tomography Michael Rabbat IEEE Infocom, Hong Kong Wednesday, March 10, 2004 Co-Authors: Mark Coates and Robert.

Canonical Subproblem: Two Senders & Two Receivers

two sender, two receiver problem characterizes network tomography problem in general

Page 7: Multiple Source, Multiple Destination Network Tomography Michael Rabbat IEEE Infocom, Hong Kong Wednesday, March 10, 2004 Co-Authors: Mark Coates and Robert.

Shared and Non-Shared Topologies

Natural dichotomy according to “model order”

5 Links2 Internal Nodes

Shared topology

8 Links4 Internal Nodes

Non-Shared topology

Page 8: Multiple Source, Multiple Destination Network Tomography Michael Rabbat IEEE Infocom, Hong Kong Wednesday, March 10, 2004 Co-Authors: Mark Coates and Robert.

Mutual Information

Shared Non-Shared

Page 9: Multiple Source, Multiple Destination Network Tomography Michael Rabbat IEEE Infocom, Hong Kong Wednesday, March 10, 2004 Co-Authors: Mark Coates and Robert.

Mutual Information

Shared Non-Shared

Same branching point Shared component links

Different branching points No shared component links

Combine Measurements!

Page 10: Multiple Source, Multiple Destination Network Tomography Michael Rabbat IEEE Infocom, Hong Kong Wednesday, March 10, 2004 Co-Authors: Mark Coates and Robert.

Arrival Order and Model Order Selection

1 1 Intuition:• Packets from A,B to 1 mix at joining point• Arrival order fixed at joining point

Assume:• Unique routes between end-hosts• Routes are stationary (5-10min) (Zhang, Paxson, Shenker, ’00)• No reordering (Bellardo & Savage, ’02)

Packets from each sender to receiver 1

Page 11: Multiple Source, Multiple Destination Network Tomography Michael Rabbat IEEE Infocom, Hong Kong Wednesday, March 10, 2004 Co-Authors: Mark Coates and Robert.

Multiple Source Active Probing

1

1

2

2

random offset

Page 12: Multiple Source, Multiple Destination Network Tomography Michael Rabbat IEEE Infocom, Hong Kong Wednesday, March 10, 2004 Co-Authors: Mark Coates and Robert.

All Packets to Receiver 1

1

1

2

2

random offset

2

1

2

1

j

repeat many times …1 = percentage different arrival order (should be very small)

Page 13: Multiple Source, Multiple Destination Network Tomography Michael Rabbat IEEE Infocom, Hong Kong Wednesday, March 10, 2004 Co-Authors: Mark Coates and Robert.

All Packets to Receiver 2

1

1

2

2

random offset

j

2

1

2

1

repeat many times …2 = percentage different arrival order (also very small)

Page 14: Multiple Source, Multiple Destination Network Tomography Michael Rabbat IEEE Infocom, Hong Kong Wednesday, March 10, 2004 Co-Authors: Mark Coates and Robert.

Send to Both Receivers

1

1

2

2

random offset

22

11 repeat many times …

percentage different arrival order(should it be small?)

Page 15: Multiple Source, Multiple Destination Network Tomography Michael Rabbat IEEE Infocom, Hong Kong Wednesday, March 10, 2004 Co-Authors: Mark Coates and Robert.

1

1

2

2

random offset

Test: Shared

Shared:

single, sharedjoining point j

Page 16: Multiple Source, Multiple Destination Network Tomography Michael Rabbat IEEE Infocom, Hong Kong Wednesday, March 10, 2004 Co-Authors: Mark Coates and Robert.

1

1

2

2

random offset

Test: Shared vs. Non-Shared

Shared: vs. Non-Shared:

multiplejoining points

j j

Page 17: Multiple Source, Multiple Destination Network Tomography Michael Rabbat IEEE Infocom, Hong Kong Wednesday, March 10, 2004 Co-Authors: Mark Coates and Robert.

Arrival Order Based Topology ID

Rice LAN

Page 18: Multiple Source, Multiple Destination Network Tomography Michael Rabbat IEEE Infocom, Hong Kong Wednesday, March 10, 2004 Co-Authors: Mark Coates and Robert.

Joint Performance & Topology Estimation

1

2

u

Performance Assessment• Link-level parameters 1, 2, …• Packet-pair measurements

1

21

21

2

Topology Characterization• Different arrival order probabilities , 1, 2

• Arrival order measurements

Page 19: Multiple Source, Multiple Destination Network Tomography Michael Rabbat IEEE Infocom, Hong Kong Wednesday, March 10, 2004 Co-Authors: Mark Coates and Robert.

Decision-Theoretic Framework

HS:

HN:Two branching, joining points unrestricted N 2 unrestricted N 2 [0,1]3

Unique joining point 2536 S 2 1=2= S 2 [0,1]1

Page 20: Multiple Source, Multiple Destination Network Tomography Michael Rabbat IEEE Infocom, Hong Kong Wednesday, March 10, 2004 Co-Authors: Mark Coates and Robert.

Characterize Topology & Performance

Generalized Likelihood Ratio Test:

Wilks’ Theorem (’38):

Under HS:

(T ! 1)

Page 21: Multiple Source, Multiple Destination Network Tomography Michael Rabbat IEEE Infocom, Hong Kong Wednesday, March 10, 2004 Co-Authors: Mark Coates and Robert.

Performance Simulation in ns

S S

R

R

R

RR

500k-10MbpsFTP and ExpOO

Page 22: Multiple Source, Multiple Destination Network Tomography Michael Rabbat IEEE Infocom, Hong Kong Wednesday, March 10, 2004 Co-Authors: Mark Coates and Robert.

Joint Topology/Performance Estimation1000 probes

Loss Only

Arrival Order Only

Arrival Order and Loss

Pro

b. C

orre

ctly

Dec

ide

Non

-Sha

red

Prob. Falsely Decide Non-Shared

Page 23: Multiple Source, Multiple Destination Network Tomography Michael Rabbat IEEE Infocom, Hong Kong Wednesday, March 10, 2004 Co-Authors: Mark Coates and Robert.

Number of Probes Used

1000

500

200

100

Pro

b. C

orre

ctly

Dec

ide

Non

-Sha

red

Prob. Falsely Decide Non-Shared

Page 24: Multiple Source, Multiple Destination Network Tomography Michael Rabbat IEEE Infocom, Hong Kong Wednesday, March 10, 2004 Co-Authors: Mark Coates and Robert.

Concluding Remarks

Combining arrival order with joint topology/performance estimation gives us an initial step towards solving this problem

www.cae.wisc.edu/[email protected]