On Evaluating a New Class of Available Bandwidth Methods Attila - - PowerPoint PPT Presentation

▶
on evaluating a new class of available bandwidth methods
SMART_READER_LITE
LIVE PREVIEW

On Evaluating a New Class of Available Bandwidth Methods Attila - - PowerPoint PPT Presentation

Jan 16, 2023 147 likes •437 views

On Evaluating a New Class of Available Bandwidth Methods Attila Psztor Darryl Veitch Pter Hga ISMA 2003 BEst Workshop 09.12.2003 San Diego, CA 1 Outline A new class of active probing methods: the Interaction class A

slide-1
SLIDE 1

San Diego, CA

09.12.2003

1

On Evaluating a New Class of Available Bandwidth Methods

Attila Pásztor Darryl Veitch Péter Hága ISMA 2003 BEst Workshop

slide-2
SLIDE 2

San Diego, CA

09.12.2003

2

Outline

  • A new class of active probing methods:

the ‘Interaction class’

  • A methodology to evaluate, compare and develop active

probing methods

  • Some results
  • Ongoing work and future plans
slide-3
SLIDE 3

San Diego, CA

09.12.2003

3

  • Two fundamental network effects identified in the

literature as the bases of existing probing techniques

– Spacing effect

  • bottleneck spacing determining the inter arrival-time to the

receiver

  • assumes probes being in the SAME busy period at the link of

interest – Accumulation effect

  • packet size dependence of service time
  • assumes probes being in DIFFERENT busy periods at all

hops

Fundamental Network Effects

slide-4
SLIDE 4

San Diego, CA

09.12.2003

4

Spacing Effect

t= p/ µ µ µ µb

slide-5
SLIDE 5

San Diego, CA

09.12.2003

5

Accumulation effect

d1 - d2 =(p1 - p2)Σ Σ Σ Σi ( 1 /µ µ µ µi )

slide-6
SLIDE 6

San Diego, CA

09.12.2003

6

Existing Bandwidth Estimation Methods

  • Packet Pair and Packet Trains

– Spacing effect based – Probes of the pattern in the same busy period

  • One Packet Based

– Accumulation effect based – Well separated, independent probes

  • Hybrid

– Based on spacing and accumulation effects – Some probes of the pattern are designated to be in the same busy period, some to be separated

slide-7
SLIDE 7

San Diego, CA

09.12.2003

7

Existing Bandwidth Estimation Methods

  • Other methods – the ‘Interaction class’

– Based on the interaction of cross traffic and probes – The cross traffic determines which probes of the pattern join the same busy period – The focus of these methods is to detect when and which probes join the same busy period, to detect the transition from an independent system to a linear system

slide-8
SLIDE 8

San Diego, CA

09.12.2003

8

Methods of the ‘Interaction’ class

  • TOPP (Melander et.al)
  • Pathload (Dovrolis et.al)
  • Pathchirp (Ribeiro et.al)
  • ‘(Poly)Chirp’
  • ‘(Poly)S(moothed)chirp’
slide-9
SLIDE 9

San Diego, CA

09.12.2003

9

Methods of the ‘Interaction’ class

  • TOPP (Melander et.al)
  • Pathload (Dovrolis et.al)
  • Pathchirp (Ribeiro et.al)
  • ‘(Poly)Chirp’
  • ‘(Poly)S(moothed)chirp’
slide-10
SLIDE 10

San Diego, CA

09.12.2003

10

Methods of the ‘Interaction’ class

  • TOPP (Melander et.al)
  • Pathload (Dovrolis et.al)
  • Pathchirp (Ribeiro et.al)
  • ‘(Poly)Chirp’
  • ‘(Poly)S(moothed)chirp’
slide-11
SLIDE 11

San Diego, CA

09.12.2003

11

Transition from an Independent System to a Linear System

slide-12
SLIDE 12

San Diego, CA

09.12.2003

12

Transition – Key to the Other Methods

slide-13
SLIDE 13

San Diego, CA

09.12.2003

13

Transition – Key to the Other Methods

slide-14
SLIDE 14

San Diego, CA

09.12.2003

14

The Key Elements of Active Probing Methods

slide-15
SLIDE 15

San Diego, CA

09.12.2003

15

Problems of Studying Active Probing Methods

  • The Network is typically

– not fully known, neither the background traffic, nor the topology – not under our control

  • Consequences:

– difficult to verify the results – hard to repeat the experiments – problematic to reliably compare different methods

slide-16
SLIDE 16

San Diego, CA

09.12.2003

16

  • Studying Active Probing Methods
  • The Simulated Network is

– fully known, including the background traffic and the topology – “fully instrumented” – fully under our control

  • As a consequence it is easy to:

– verify the results – repeat the experiments – reliably compare different methods

  • Essential to ensure, that the

topology and background traffic is realistic

slide-17
SLIDE 17

San Diego, CA

09.12.2003

17

PSIM - The Probing Simulator

– High performance, tailored to the requirements of active probing – A multiple hop network model – series of FIFO queues and links – Inputs:

  • Route configuration
  • Cross traffic entering at different hops
  • Probe stream injected

– The output is the probe arrival time series to the intermediate hops and to the receiver, including information about probes being I/B

  • Calculated using the recursion relations of queueing theory

– Accepts arbitrary cross traffic – generated synthetically or from real network traces – Compatible input/output formats with our active probing applications

  • Allowing easy switching between simulated and real network

experiments

slide-18
SLIDE 18

San Diego, CA

09.12.2003

18

An experiment using PSIM

  • We have integrated pathload into our framework
  • Compare it to our chirp pattern based AB estimation

method

  • A two hop route - a100Mbps link followed by a 3Mbps link
  • A 30 min long cross traffic trace
  • Pathload is probing with 220 kbps average rate
  • The chirp based method is probing with 42 kbps rate
slide-19
SLIDE 19

San Diego, CA

09.12.2003

19

slide-20
SLIDE 20

San Diego, CA

09.12.2003

20

slide-21
SLIDE 21

San Diego, CA

09.12.2003

21

slide-22
SLIDE 22

San Diego, CA

09.12.2003

22

slide-23
SLIDE 23

San Diego, CA

09.12.2003

23

slide-24
SLIDE 24

San Diego, CA

09.12.2003

24

slide-25
SLIDE 25

San Diego, CA

09.12.2003

25

slide-26
SLIDE 26

San Diego, CA

09.12.2003

26

The proof of the pudding …

  • The simulation environment is not identical to the reality

– SW and HW limitations of the probing equipment – Some properties of real networks may differ from the simulated model

  • Ongoing and future work:

– Integrating pathchirp into our experimental framework – Study of the properties of the ‘interaction class’ – Development of a new standalone available bandwidth estimation application

slide-27
SLIDE 27

San Diego, CA

09.12.2003

27

Chirp based probing