PRICING FOR MOBILE DATA Sangtae Ha Princeton University Joint work - - PowerPoint PPT Presentation

▶
pricing for mobile data
SMART_READER_LITE
LIVE PREVIEW

PRICING FOR MOBILE DATA Sangtae Ha Princeton University Joint work - - PowerPoint PPT Presentation

Sep 19, 2022 126 likes •671 views

TUBE: TIME DEPENDENT PRICING FOR MOBILE DATA Sangtae Ha Princeton University Joint work with: Soumya Sen, Carlee Joe-Wong, Youngbin Im, Mung Chiang 2 (Mobile) Data Explosion Mobile data growing at 78% annually 2.4 billion mobile users

slide-1
SLIDE 1

TUBE: TIME DEPENDENT PRICING FOR MOBILE DATA

Sangtae Ha Princeton University Joint work with: Soumya Sen, Carlee Joe-Wong, Youngbin Im, Mung Chiang

slide-2
SLIDE 2

(Mobile) Data Explosion

Mobile data growing at 78% annually 2.4 billion mobile users worldwide, 260 million US mobile data users by 2017

2

slide-3
SLIDE 3

Driving Forces

3

Mobile Video Cloud Sync Data-hungry Apps High-res Devices A Perfect Storm

slide-4
SLIDE 4

Industry Moves: US ISPs

4

AT&T throttles unlimited iPhone users (July 2011) 9 AT&T starts $10/GB

  • verage fee for

smartphone data plans (June 2010) 6 Comcast caps data at 250 GB (August 2008)2

Wireless Wireline

Introduction of Data Caps

Verizon eliminates new unlimited smartphone plans (July 2011) 8

Elimination of Unlimited Data Plans Introduction of Usage-based penalties

T-Mobile starts data caps and throttling penalty (May 2011) 7 AT&T caps U-verse to 250 GB and DSL to 150 GB with a $10 penalty for an additional 50 GB (May 2011) 4 Time-Warner trials data caps in Texas (June 2008)1 AT&T starts throttling wireline users (April 2011) 3 Verizon AT&T introduce shared data plans with unlimited voice and text (June 2012) 12 Verizon to eliminate unlimited data plans (May 2012) 11 No unlimited plans

  • ffered for iPad LTE

(March 2012) 10 Comcast moves towards tiered usage- based billing (May 2012) 5

slide-5
SLIDE 5

But Not Heavy All the Time

5

How to leverage the peak-valley differential? Average demand < 30% Peak demand > 99%

slide-6
SLIDE 6

6

slide-7
SLIDE 7

7

slide-8
SLIDE 8

8

slide-9
SLIDE 9

Opportunities

Time Elasticity: Opportunities

9

Time Elasticity

Volume

Streaming videos, Gaming Texting, Weather, Finance Email, Social Network updates Software Downloads Movies & Multimedia downloads, P2P Cloud

Pricing Survey: http://arxiv.org/abs/1201.4197

slide-10
SLIDE 10

10

slide-11
SLIDE 11

11

slide-12
SLIDE 12

12

slide-13
SLIDE 13

13

slide-14
SLIDE 14

14

slide-15
SLIDE 15

15

slide-16
SLIDE 16

16

slide-17
SLIDE 17

17

slide-18
SLIDE 18

18

slide-19
SLIDE 19

Contributions

  • 1. An architecture and a fully functional

system for offering TDP for mobile data

  • 1. User behavior models and optimized

price computation

  • 1. A realistic evaluation with real users

19

slide-20
SLIDE 20

TDP Overview

20

Internet Prices Traffic

slide-21
SLIDE 21

TUBE Theory

21

slide-22
SLIDE 22

Feedback Loop

22

Prices Usage User Interface Price Calculation Network Measurement User Behavior Estimation

slide-23
SLIDE 23

Waiting Functions

23

  • Probabilistically estimate “willingness” to wait

r1 r2 r3

= impatience

r m1wr1 +m2wr2 +m3wr3

Estimate parameters

w

discount time waited

slide-24
SLIDE 24

Minimizing Cost

24

Time Traffic

Offering discounts Exceeding capacity

G1 +G2

minimize variables di

slide-25
SLIDE 25

25

slide-26
SLIDE 26

26

slide-27
SLIDE 27

27

slide-28
SLIDE 28

TUBE Architecture

28

slide-29
SLIDE 29

Design Guidelines

  • 1. Separating functionality

 Price computation on a central server  Price display and scheduling the usage on the user devices

  • 2. Scaling up the system

 User behavior estimation algorithm requires only aggregate,

not individual usage data

 Formulate the price calculation as a convex optimization for

computation scalability for many TDP periods

  • 3. Protecting user privacy

 No Deep Packet Inspection (DPI) and no private data is exchanged.

  • 4. Empowering user control

29

slide-30
SLIDE 30

TUBE: TDP Architecture

30

User Device

Autopilot

Youtube Flipboard Magazine Netflix Apple App Store

Usage Monitor

User GUI ISP Server Aggregate Traffic Measurement

Allow or Block

Price Information Application Traffic Secure Connection

User Behavior Estimation Price

  • ptimizer

! "#$% & ' (% ) * +, - . & / 00& 1+2% ) 34 % (&

slide-31
SLIDE 31

TUBE Implementation

31

slide-32
SLIDE 32

TDP Price Optimizer Pricing Plans

!

Manual Autopilot Pricing Policy Container Authentication Module Delegation Mechanism

User Device Queries

REST API

User Behavior Estimation Measurement Usage Monitoring Plugins Linux Netfilter

3G Standard

RADIUS

!

Pricing Policy Enforcer Plugins Linux Traffic Control DPI TDP

3G Standard

Linux Netfilter Push Notifier

SMS

Push Message

Server Side Design

32

slide-33
SLIDE 33

TDP Price Optimizer Pricing Plans

!

Manual Autopilot Pricing Policy Container Authentication Module Delegation Mechanism

User Device Queries

REST API

User Behavior Estimation Measurement Usage Monitoring Plugins Linux Netfilter

3G Standard

RADIUS

!

Pricing Policy Enforcer Plugins Linux Traffic Control DPI TDP

3G Standard

Linux Netfilter Push Notifier

SMS

Push Message

Server Side Design

33

Number of Periods 12 24 48 96 144 Behavior Estimation (sec) 12.76 200.0 959.6 1967 15040 Price Calculation (sec) 1.67 1.69 1.70 1.81 1.84

Number of TDP periods

Number of Application Type Number of Periods 2 4 8 12 0.21 12.99 21.52 24 3.33 47.08 75.47 48 15.99 197.22 215.42

Number of TDP periods x Number of application types (mins)

slide-34
SLIDE 34

Client Side Design

34

Manual Autopilot Communication Module Graphical User Interface

Budget

Price Dispatcher Session Recorder Usage Tracker Price Display Usage Display Settings Display Top 5 Apps Display Current Bill Display Popup Display Price Status bar Display Price DB Daemon Usage Collector Usage DB Budget Helper

App Scheduler

Budget Manager Delegation Pulled Local Autopilot Algorithm

App PPI Scheduler

Enforcer Allow/Block Notifier Task Manager

Type Status bar App usage Daemon Support LOC iPhone No No Partial 25K Android Yes Yes Yes 5.4K Windows Yes Yes Yes 5.3K

slide-35
SLIDE 35

TUBE Princeton Trial

35

slide-36
SLIDE 36

Princeton Trial: Money Flow

  • 50 AT&T participants: 27 iPhones, 23 iPads
  • Faculty, staff, and students
  • 14 Academic Departments & Divisions

36

Participants

$$ $$

TUBE Project

$$

TDP based payments Current pricing scheme

Wireless Provider

AT&T

slide-37
SLIDE 37

Princeton Trial: Data Flow

37

User's iPhone, iPad

AT&T's mobile network DNS NAT Gateway AT&T Firewall TUBE Servers

VLR MCS SGSN AuC GGSN HLR

3G Core Network

GMSC

PSTN Data Flow

BSS

Data Flow VPN

BSC

slide-38
SLIDE 38

TUBE App: Information Screens

38

4 % 4 % 4 %

Price indicator

slide-39
SLIDE 39

TUBE App: Scheduling Screens

39

slide-40
SLIDE 40

Princeton Trial Results

40

slide-41
SLIDE 41

Two main goals of the trial

  • 1. How do people respond to pricing changes and

GUI design?

  • 1. Can the end to end TDP system work in the

real world and can our architecture scale up?

41

slide-42
SLIDE 42

Usage Statistics

  • How much bandwidth participants use? – „Heavy tailed‟
  • Which applications use the most bandwidth – Streaming and

surfing

42 0.1 0.2 0.3 0.4 0.5 Movie Web Downloads Music News/Mags. Other

Usage Volume (GB) Applica on Type

Non-Jailbroken iPads Jailbroken iPads iPhones

slide-43
SLIDE 43

UI Effectiveness

43

Do users respond more to the numerical values of TDP prices

  • r to the color of the price indicator bar on the home screen?

 Users paid more attention to indicator color than the numerical discount values

slide-44
SLIDE 44

UI Effectiveness

Period types 1 and 3

44

 Users paid more attention to indicator color than the numerical discount values

Period types 2 and 1

Type Periods First Stage Color Discount Second Stage Color Discount

1 2, 8, 14, 20 Orange 10% Orange 28% 2 3, 6, …, 24 Orange 10% Green 30% 3 5, 11, 17, 23 Orange 10% Orange 9%

slide-45
SLIDE 45

Optimized TDP Impact

45

Does the peak usage decrease with time-dependent pricing? And does this decrease come at the expense of an overall decrease in usage?

» Optimized TDP reduces the peak-to-average ratio » Overall usage significantly increases with TDP

slide-46
SLIDE 46

Optimized TDP Impact

46

Does the peak usage decrease with time-dependent pricing? And does this decrease come at the expense of an overall decrease in usage?

» Optimized TDP reduces the peak-to-average ratio » Overall usage significantly increases with TDP

PAR reduces by 30% Overall usage increases by 130%

slide-47
SLIDE 47

Trial Limitations and Extensions

Limitations:

  • 1. Single bottleneck
  • 2. Mobility
  • 3. Control group
  • 4. Time granularity

Extensions:

  • 1. Location/congestion dependent pricing
  • 2. Commercial operator trials

47

slide-48
SLIDE 48

Summary

  • 1. A fully functional system for offering TDP for

mobile data

  • 2. People are sensitive to time-dependent

prices and indeed shift their Internet usage to off-peak periods

  • 3. The pilot trial motivates future study on TDP

for different markets and demographics

48

slide-49
SLIDE 49

Thank you!

sangtaeh@princeton.edu

 DataMi: http://www.datami.com  DataWiz: http://www.datawizapp.com  SDP Forum: http://scenic.princeton.edu/SDP2012

49

BE TU

slide-50
SLIDE 50

Backup Slides

50

slide-51
SLIDE 51

TDP Performance

51

slide-52
SLIDE 52

Price Sensitivity

52

Do users wait to use mobile data in return for a monetary discount?

» Average usage decrease in high-price periods relative to the changes in low-price periods

slide-53
SLIDE 53

Notification Effectiveness

53

Do notifications impact usage?

» 80-90% of users decrease or did not increase their usage after the 1st notification » For all subsequent notifications, about 60-80% of the active users decrease their usage, while the others remained price-insensitivity

slide-54
SLIDE 54

Impact on Ecosystem

54

Does the application usage distribution change due to TDP?

» People are motivated to use more bandwidth during low-price periods, “valley filling”.

0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8 Movies Web Downloads Music News/ Mags. Other

Usage (%

  • f

Total) Applica on Type

Before TDP With TDP