traffic flow estimation from cellular network data Masters thesis - - PowerPoint PPT Presentation

▶
traffic flow estimation from cellular network data
SMART_READER_LITE
LIVE PREVIEW

traffic flow estimation from cellular network data Masters thesis - - PowerPoint PPT Presentation

Mar 11, 2023 251 likes •790 views

traffic flow estimation from cellular network data Masters thesis presentation Nils Breyer August 27, 2015 LiU/ITN, UC Berkeley (USA) why are linkflows of interest? Linkflow number of vehicles or people using a link during a certain time

slide-1
SLIDE 1

traffic flow estimation from cellular network data

Master’s thesis presentation

Nils Breyer August 27, 2015

LiU/ITN, UC Berkeley (USA)

slide-2
SLIDE 2

why are linkflows of interest?

Linkflow number of vehicles or people using a link during a certain time Usage

  • Traffic planning
  • Traffic control

2

slide-3
SLIDE 3

why are linkflows of interest?

Linkflow number of vehicles or people using a link during a certain time Usage

  • Traffic planning
  • Traffic control

2

slide-4
SLIDE 4

datasources

Travel surveys Link counts Cellular networks

3

slide-5
SLIDE 5

datasources

Travel surveys Link counts Cellular networks

3

slide-6
SLIDE 6
  • utline
  • Overview of the model
  • Trip extraction and travel demand estimation
  • Cellpath routing
  • Network loading
  • Results & validation

4

slide-7
SLIDE 7

Overview of the model

5

slide-8
SLIDE 8

cellpath

Cellpath a sequence of cells that a user connected to along a trip

6

slide-9
SLIDE 9
  • verview of the model

Cellpath routing

Celltower locations

Network loading

Route link counts Cellular data 7

slide-10
SLIDE 10
  • verview of the model

Cellpath routing

Celltower locations

Network loading

Route link counts Cellular data 7

slide-11
SLIDE 11

problem: data only for a sample

8

slide-12
SLIDE 12

problem: data only for a sample

8

slide-13
SLIDE 13
  • d-matrix

OD-matrix A matrix containing the traffic demand between pairs of origin and destination zones

9

slide-14
SLIDE 14
  • verview of the model

OD matrix estimation Cellpath routing

Celltower locations

Network loading

Route link counts Time-sliced OD matrix Cellular data

10

slide-15
SLIDE 15
  • verview of the model

OD matrix estimation Cellpath routing

Celltower locations

Network loading

Route link counts Time-sliced OD matrix Cellular data

10

slide-16
SLIDE 16

problem: data contains not only movements

User ID Timestamp Cell ID 1 2013-10-01 06:50:00 1 1 2013-10-01 08:10:00 1 1 2013-10-01 08:10:00 3 2 2013-10-01 08:20:00 2

Table 1: An example of a cellular network dataset

Datatypes

  • Call detail records (CDR)
  • Handover data

11

slide-17
SLIDE 17

problem: data contains not only movements

User ID Timestamp Cell ID 1 2013-10-01 06:50:00 1 1 2013-10-01 08:10:00 1 1 2013-10-01 08:10:00 3 2 2013-10-01 08:20:00 2

Table 1: An example of a cellular network dataset

Datatypes

  • Call detail records (CDR)
  • Handover data

11

slide-18
SLIDE 18
  • verview of the model

OD matrix estimation Cellpath routing

Celltower locations

Network loading

Linkpaths link counts Time-sliced OD matrix Trips with cellpath

Trip extraction

Cellular data

12

slide-19
SLIDE 19

Trip extraction & travel demand estimation

13

slide-20
SLIDE 20
  • verview of the model

OD matrix estimation Cellpath routing

Celltower locations

Network loading

Linkpaths link counts Time-sliced OD matrix Trips with cellpath

Trip extraction

Cellular data

14

slide-21
SLIDE 21

trip extraction

Challenges

  • Depends on the datasource
  • Cell-switching due to network balancing
  • Low resolution in time (CDR)

Solutions

  • algorithmic approach (CDR)
  • movement efficiency metric (handover data)

15

slide-22
SLIDE 22

trip extraction

Challenges

  • Depends on the datasource
  • Cell-switching due to network balancing
  • Low resolution in time (CDR)

Solutions

  • algorithmic approach (CDR)
  • movement efficiency metric (handover data)

15

slide-23
SLIDE 23

travel demand estimation

Challenges

  • Additional data necessary
  • Sample might not be representative
  • Vehicles ̸= people

Solutions

  • Upscaling using total population
  • Data fusion with census data

16

slide-24
SLIDE 24

travel demand estimation

Challenges

  • Additional data necessary
  • Sample might not be representative
  • Vehicles ̸= people

Solutions

  • Upscaling using total population
  • Data fusion with census data

16

slide-25
SLIDE 25

Cellpath routing

17

slide-26
SLIDE 26
  • verview of the model

OD matrix estimation Cellpath routing

Celltower locations

Network loading

Linkpaths link counts Time-sliced OD matrix Trips with cellpath

Trip extraction

Cellular data

18

slide-27
SLIDE 27

cellpath routing problem

Input Cellpath p = (b1, b2, ..., bn) Output A route consisting of connected links on the road network Goal Recover the original route that the user took

19

slide-28
SLIDE 28

voronoi diagram

20

slide-29
SLIDE 29

cellpath routing algorithms

Shortest-path Routing shortest path between first and last cell in cellpath Strict Voronoi Routing route must go through every cell in the cellpath Lazy Voronoi Routing route is encouraged to go through cells in the cellpath

21

slide-30
SLIDE 30

shortest-path routing

shortest path between first and last cell in cellpath

22

slide-31
SLIDE 31

strict voronoi routing

route must go through every cell in the cellpath

23

slide-32
SLIDE 32

waypoint search

3 2 1

Antenna position Boundary junction (possible waypoint)

A B C

24

slide-33
SLIDE 33

lazy voronoi routing

route is encouraged to go through cells in the cellpath

25

slide-34
SLIDE 34

comparision between algorithms

Shortest-path Strict Voronoi Lazy Voronoi

26

slide-35
SLIDE 35

Network loading

27

slide-36
SLIDE 36
  • verview of the model

OD matrix estimation Cellpath routing

Celltower locations

Network loading

Linkpaths link counts Time-sliced OD matrix Trips with cellpath

Trip extraction

Cellular data

28

slide-37
SLIDE 37

single od-pair

Shortest-path Strict Voronoi Lazy Voronoi

29

slide-38
SLIDE 38

single od-pair

Shortest-path Strict Voronoi Lazy Voronoi

29

slide-39
SLIDE 39

single od-pair

Shortest-path Strict Voronoi Lazy Voronoi

29

slide-40
SLIDE 40

Results & validation

30

slide-41
SLIDE 41

los angeles dataset

Coverage I-210 corridor in Los Angeles, USA Input data

  • Antenna positions
  • OD-matrix estimated from real cellular data (UC Berkeley)
  • Simulated cellpaths (UC Berkeley)
  • OpenStreetMap road network

31

slide-42
SLIDE 42

route validation

Cellpath routing

Celltower locations

Validation

cellpath estimated route MATSim route

MatSim simulation

OD-matrix

Definition: Route similarity metric Two routes ra Va Ea and rb Vb Eb have a route similarity of: s a b Va Vb Va Vb

32

slide-43
SLIDE 43

route validation

Cellpath routing

Celltower locations

Validation

cellpath estimated route MATSim route

MatSim simulation

OD-matrix

Definition: Route similarity metric Two routes ra = (Va, Ea) and rb = (Vb, Eb) have a route similarity of: s(a, b) := |Va ∩ Vb| |Va ∪ Vb|

32

slide-44
SLIDE 44

route validation

0.25) 0.31) 0.43) E0.20) 0.00) 0.20) 0.40) 0.60) 0.80) 1.00) Similarity2 Shortest) Strict) Lazy)(ds)=)0.03))

average of 1000 randomly selected routes

33

slide-45
SLIDE 45

network loading result

using Lazy Voronoi routing

34

slide-46
SLIDE 46

network loading validation

35

slide-47
SLIDE 47

Conclusions

36

slide-48
SLIDE 48

conclusions

  • 1. Lazy Voronoi improves route recoverage over Shortest-path and

Strict Voronoi routing

  • 2. Route similarity not enough for a precise network loading
  • 3. Data fusion with traffic counts could improve the result

37

slide-49
SLIDE 49

Nils Breyer

nilsbreyer.eu

38

slide-50
SLIDE 50

4-stage model

Trip distribution Mode choice

Route choice & network loading

link counts

Trip extraction 39

slide-51
SLIDE 51

line simplification

1 2 3 4

Ramer-Douglas Peuker algorithm

40

slide-52
SLIDE 52

lazy voronoi routing details

¡

S ¡ W ¡

¡

¡ ¡

T ¡

S ¡ T ¡

Voronoi-­‑routed ¡path ¡ Actual ¡path ¡ ¡ Cellpath ¡ Simplified ¡cellpath ¡ Lazy ¡Voronoi ¡route ¡ ¡

41

slide-53
SLIDE 53

la network loading geh

42