palm calculus part 1 the importance of the viewpoint jy le boudec 1 may 2015
DESCRIPTION
Example: Gatekeeper; Average execution time Real time t (ms) job arrival Execution time for a job that arrives at t (ms) Viewpoint 1: System Designer Viewpoint 2: CustomerTRANSCRIPT
![Page 1: Palm Calculus Part 1 The Importance of the Viewpoint JY Le Boudec 1 May 2015](https://reader036.vdocument.in/reader036/viewer/2022062905/5a4d1ae87f8b9ab059979c87/html5/thumbnails/1.jpg)
Palm CalculusPart 1
The Importance of the Viewpoint
JY Le Boudec
1May 2015
![Page 2: Palm Calculus Part 1 The Importance of the Viewpoint JY Le Boudec 1 May 2015](https://reader036.vdocument.in/reader036/viewer/2022062905/5a4d1ae87f8b9ab059979c87/html5/thumbnails/2.jpg)
1. Event versus Time AveragesConsider a simulation, state St
Assume simulation has a stationary regime
Consider an Event Clock: times Tn at which some specific changes of state occur
Ex: arrival of job; Ex. queue becomes empty
Event average statistic
Time average statistic
2
![Page 3: Palm Calculus Part 1 The Importance of the Viewpoint JY Le Boudec 1 May 2015](https://reader036.vdocument.in/reader036/viewer/2022062905/5a4d1ae87f8b9ab059979c87/html5/thumbnails/3.jpg)
Example: Gatekeeper; Average execution time
3
0 90 100 190 200 290 300
50001000
Real time t (ms)
job arrival
50001000
50001000
Execution time for a job that arrives at t (ms)
Viewpoint 1: System Designer Viewpoint 2: Customer
![Page 4: Palm Calculus Part 1 The Importance of the Viewpoint JY Le Boudec 1 May 2015](https://reader036.vdocument.in/reader036/viewer/2022062905/5a4d1ae87f8b9ab059979c87/html5/thumbnails/4.jpg)
Sampling Bias
Ws and Wc are different
A metric definition should mention the sampling method (viewpoint)Different sampling methods may provide different values: this is the sampling bias
Palm Calculus is a set of formulas for relating different viewpoints
Can often be obtained by means of the Large Time Heuristic
5
![Page 5: Palm Calculus Part 1 The Importance of the Viewpoint JY Le Boudec 1 May 2015](https://reader036.vdocument.in/reader036/viewer/2022062905/5a4d1ae87f8b9ab059979c87/html5/thumbnails/5.jpg)
Large Time Heuristic Explained
on an Example
We want to relate and We apply the large time heuristic
1. How do we evaluate these metrics in a simulation ?
where index of next green or red arrow at or after
6
𝑋 1𝑋 2𝑋 3𝑋 3
𝑋 5𝑋 6
𝑆3
![Page 6: Palm Calculus Part 1 The Importance of the Viewpoint JY Le Boudec 1 May 2015](https://reader036.vdocument.in/reader036/viewer/2022062905/5a4d1ae87f8b9ab059979c87/html5/thumbnails/6.jpg)
Large Time Heuristic Explained
on an Example
2. Break one integral into pieces that match the ’s:
7
𝑋 1𝑋 2𝑋 3𝑋 3
𝑋 5𝑋 6
𝑆3
![Page 7: Palm Calculus Part 1 The Importance of the Viewpoint JY Le Boudec 1 May 2015](https://reader036.vdocument.in/reader036/viewer/2022062905/5a4d1ae87f8b9ab059979c87/html5/thumbnails/7.jpg)
Large Time Heuristic Explained
on an Example
3. Compare
9
𝑋 1𝑋 2𝑋 3𝑋 3
𝑋 5𝑋 6
𝑆3
![Page 8: Palm Calculus Part 1 The Importance of the Viewpoint JY Le Boudec 1 May 2015](https://reader036.vdocument.in/reader036/viewer/2022062905/5a4d1ae87f8b9ab059979c87/html5/thumbnails/8.jpg)
This is Palm Calculus !
11
𝑊 𝑐=𝜆cov (𝑆 ,𝑋 )+𝑊 𝑠
𝑺𝒏
𝑿𝒏
![Page 9: Palm Calculus Part 1 The Importance of the Viewpoint JY Le Boudec 1 May 2015](https://reader036.vdocument.in/reader036/viewer/2022062905/5a4d1ae87f8b9ab059979c87/html5/thumbnails/9.jpg)
In which case do we expect to see
A. Sn = 90, 10, 90, 10, 90; Xn = 5000, 1000, 5000, 1000, 5000B. Sn = 90, 10, 90, 10, 90; Xn = 1000, 5000, 1000, 5000, 1000
C. BothD. NoneE. I don’t know
12
72%
0%0%
28%
0%
![Page 10: Palm Calculus Part 1 The Importance of the Viewpoint JY Le Boudec 1 May 2015](https://reader036.vdocument.in/reader036/viewer/2022062905/5a4d1ae87f8b9ab059979c87/html5/thumbnails/10.jpg)
The Large Time Heuristic
Formally correct if simulation is stationary
It is a robust method, i.e. independent of assumptions on distributions (and on independence)
14
![Page 11: Palm Calculus Part 1 The Importance of the Viewpoint JY Le Boudec 1 May 2015](https://reader036.vdocument.in/reader036/viewer/2022062905/5a4d1ae87f8b9ab059979c87/html5/thumbnails/11.jpg)
Other «Clocks»
15
Flow 1 Flow 2
Flow 3
Distribution of flow sizesfor an arbitrary flowfor an arbitrary packet
![Page 12: Palm Calculus Part 1 The Importance of the Viewpoint JY Le Boudec 1 May 2015](https://reader036.vdocument.in/reader036/viewer/2022062905/5a4d1ae87f8b9ab059979c87/html5/thumbnails/12.jpg)
Which curves are for the per-packet
viewpoint ?
A. AB. BC. It dependsD. I don’t know
16
71%
18%
0%
12%
![Page 13: Palm Calculus Part 1 The Importance of the Viewpoint JY Le Boudec 1 May 2015](https://reader036.vdocument.in/reader036/viewer/2022062905/5a4d1ae87f8b9ab059979c87/html5/thumbnails/13.jpg)
Mean flow size:per flow per packet
19
Flow 1 Flow 2
Flow 3
Distribution of flow sizesfor an arbitrary flowfor an arbitrary packet
![Page 14: Palm Calculus Part 1 The Importance of the Viewpoint JY Le Boudec 1 May 2015](https://reader036.vdocument.in/reader036/viewer/2022062905/5a4d1ae87f8b9ab059979c87/html5/thumbnails/14.jpg)
Large «Time» Heuristic1. How do we evaluate these metrics in a simulation ?per flow per packet where when packet belongs to flow
20
![Page 15: Palm Calculus Part 1 The Importance of the Viewpoint JY Le Boudec 1 May 2015](https://reader036.vdocument.in/reader036/viewer/2022062905/5a4d1ae87f8b9ab059979c87/html5/thumbnails/15.jpg)
Large «Time» Heuristic1. How do we evaluate these metrics in a simulation ?
per flow per packet where when packet belongs to flow
2. Put the packets side by side, sorted by flow
21
Flow n=1 Flow n=2 Flow n=3
p=1 p=2 p=3 p=4 p=5 p=6 p=7 p=8 p=9
𝑆𝑃=1𝑃 (𝑆1+𝑆1+𝑆2+𝑆2+𝑆3+𝑆3+𝑆3+𝑆3+𝑆3+…)
Size Size Size
![Page 16: Palm Calculus Part 1 The Importance of the Viewpoint JY Le Boudec 1 May 2015](https://reader036.vdocument.in/reader036/viewer/2022062905/5a4d1ae87f8b9ab059979c87/html5/thumbnails/16.jpg)
Large «Time» Heuristic
3. Compare
22
Flow n=1 Flow n=2 Flow n=3
p=1 p=2 p=3 p=4 p=5 p=6 p=7 p=8 p=9
Size Size Size
![Page 17: Palm Calculus Part 1 The Importance of the Viewpoint JY Le Boudec 1 May 2015](https://reader036.vdocument.in/reader036/viewer/2022062905/5a4d1ae87f8b9ab059979c87/html5/thumbnails/17.jpg)
Large «Time» Heuristic for PDFs of flow sizesPut the packets side by side, sorted by flow
1. How do we evaluate these metrics in a simulation ?
23
Flow n=1 Flow n=2 Flow n=3
![Page 18: Palm Calculus Part 1 The Importance of the Viewpoint JY Le Boudec 1 May 2015](https://reader036.vdocument.in/reader036/viewer/2022062905/5a4d1ae87f8b9ab059979c87/html5/thumbnails/18.jpg)
24
![Page 19: Palm Calculus Part 1 The Importance of the Viewpoint JY Le Boudec 1 May 2015](https://reader036.vdocument.in/reader036/viewer/2022062905/5a4d1ae87f8b9ab059979c87/html5/thumbnails/19.jpg)
25
![Page 20: Palm Calculus Part 1 The Importance of the Viewpoint JY Le Boudec 1 May 2015](https://reader036.vdocument.in/reader036/viewer/2022062905/5a4d1ae87f8b9ab059979c87/html5/thumbnails/20.jpg)
Cyclist’s Paradox
On a round trip tour, there is more uphill than downhill
26
![Page 21: Palm Calculus Part 1 The Importance of the Viewpoint JY Le Boudec 1 May 2015](https://reader036.vdocument.in/reader036/viewer/2022062905/5a4d1ae87f8b9ab059979c87/html5/thumbnails/21.jpg)
The km clock vs the standard clock
speed for the kilometer
27
![Page 22: Palm Calculus Part 1 The Importance of the Viewpoint JY Le Boudec 1 May 2015](https://reader036.vdocument.in/reader036/viewer/2022062905/5a4d1ae87f8b9ab059979c87/html5/thumbnails/22.jpg)
BorduRail claims that only 5% of trains arrivals are late
BorduKonsum claims that 30% of train users suffer from late train
arrivals
A. At least one of them liesB. The number of passengers in a
late train passengers in average train
C. The number of passengers in a late train passengers in average train
D. I don’t know
28
0% 0%
44%
56%