Televisions, Video Privacy, and Powerline Electromagnetic - - PowerPoint PPT Presentation

▶
televisions video privacy and powerline
SMART_READER_LITE
LIVE PREVIEW

Televisions, Video Privacy, and Powerline Electromagnetic - - PowerPoint PPT Presentation

Jun 22, 2023 580 likes •1.23k views

Televisions, Video Privacy, and Powerline Electromagnetic Interference Miro Enev , Sidhant Gupta, Yoshi Kohno, & Shwetak Patel Security & UbiComp Labs @ UW Smart Home = Smart Devices + Smart Sensors The Picture Today Utility power

slide-1
SLIDE 1

Televisions, Video Privacy, and Powerline Electromagnetic Interference

Miro Enev, Sidhant Gupta, Yoshi Kohno, & Shwetak Patel

Security & UbiComp Labs @ UW

slide-2
SLIDE 2

Smart Home = Smart Devices + Smart Sensors

slide-3
SLIDE 3

The Picture Today

Utility 3rd Parties power sensor

slide-4
SLIDE 4

The Picture Today

Utility 3rd Parties UbiComp power sensor

slide-5
SLIDE 5

Sophisticated electrical sensors are already deployed in homes to help achieve new efficiency and utility goals.

Emerging Trends:

Benefactors:

slide-6
SLIDE 6

What private information is available from the powerline?

Research Q:

slide-7
SLIDE 7

What private information is available from the powerline?

Research Q:

Answer:

Modern TVs leak substantial information on the power line that is indicative of the screen content

slide-8
SLIDE 8

What private information is available from the powerline?

Research Q:

Answer:

Modern TVs leak substantial information on the power line that is indicative of the screen content Information can be collected using a single sensor installed anywhere on power line

slide-9
SLIDE 9

U.S. Video Protection & Privacy Act of 1998 states that video viewing records be kept private

What’s the BIG deal?

slide-10
SLIDE 10

U.S. Video Protection & Privacy Act of 1998 states that video viewing records be kept private

What’s the BIG deal?

We hope our work can inform future discourse about the directions of Powerline sensing technologies

slide-11
SLIDE 11

Smart Sensor Power measurements reveal private information about homeowner’s activity [Markham 2010]; concurrent work also looks at TVs [Greveler 2011] In-line power measurement of a PC can reveal web browsing habits [Clark 2011]

slide-12
SLIDE 12

Power Based Measurement

In-line with device

slide-13
SLIDE 13

Voltage Based Measurement

Parallel with device

...

Home’s Powerline Infrastructure

slide-14
SLIDE 14

Switched Mode Power Supply

slide-15
SLIDE 15

Switching Circuits generate high frequency Electro Magnetic Interference (EMI)

Supply

Load Inductor

slide-16
SLIDE 16

Switching Circuits generate high frequency Electro Magnetic Interference (EMI)

Supply

Load Inductor

slide-17
SLIDE 17

Switching Circuits generate high frequency EMI

which couples onto the powerline

Supply

Load Inductor Home’s Power Line Infrastructure

slide-18
SLIDE 18

Supply

Load Inductor Home’s Power Line Infrastructure

we record the EMI

Switching Circuits generate high frequency EMI

which couples onto the powerline

slide-19
SLIDE 19

EMI: Summary

Many modern devices produce EMI! EMI can be collected with a single sensor! Signals live in the frequency domain!

Frequency (kHz)

ElectriSense (2010 Gupta et al. )

slide-20
SLIDE 20

EMI @ Home

slide-21
SLIDE 21

Movie Link

Video TV EMI

slide-22
SLIDE 22

Our 8 TVs

58” 42” 32”

Panasonic 42-A Panasonic 42-B Sharp 42 Sharp 32-A Sharp 32-B

Samsung 32

Samsung 58-A Samsung 58-B

Plasma, 2010 LCD, 2007 - 2009

slide-23
SLIDE 23

Q1: Do TVs produce repeatable EMI given repeated screen content? Q2: For a given TV, does different screen content produce different EMI? Q3: Is EMI consistent across TVs from the same model family? Q4: Can we use EMI to determine what is being watched on TV? Q5: Can we match lab EMI to EMI recorded in various home settings? Research Questions?

slide-24
SLIDE 24

Q1: Do TVs produce repeatable EMI given repeated screen content? Q2: For a given TV, does different screen content produce different EMI? Q3: Is EMI consistent across TVs from the same model family? Q4: Can we use EMI to determine what is being watched on TV? Q5: Can we match lab EMI to EMI recorded in various home settings? Research Questions?

slide-25
SLIDE 25

Screen Content = 20 IMDB Top Grossing Movies

Action Animation Comedy Documentary Drama

slide-26
SLIDE 26

Lab Setup

(I)Isolation transformer

  • external EMI filter

(A)Power Line Interface

  • custom voltage sensor

(U) Universal Software Radio Peripheral

  • analog to digital convertor

(A)Spectrum analyzer

  • visualization & logging
slide-27
SLIDE 27

Signal Extraction

Frequency (kHz)

30 100 200

Time (secs)

slide-28
SLIDE 28

Signal Extraction

Frequency (kHz)

30 100 200

Time (secs)

slide-29
SLIDE 29

Extracting TV EMI

slide-30
SLIDE 30

Extracting TV EMI

slide-31
SLIDE 31

Q1: Do TVs produce repeatable EMI given repeated screen content? Q2: For a given TV, does different screen content produce different EMI? Q3: Is EMI consistent across TVs from the same model family? Research Questions?

slide-32
SLIDE 32

Q4:

Can we use EMI to determine what is

being watched on a TV?

slide-33
SLIDE 33

Match EMI to a DB Approach:

Q4:

Can we use EMI to determine what is

being watched on a TV?

slide-34
SLIDE 34

. . .

Time (mins.) 60

. . .

Building an EMI Database

EMI

slide-35
SLIDE 35

Time (mins.) 60

Multiple TVs Same Content

Pan42A Pan42B

Multiple Runs

slide-36
SLIDE 36

Multiple Runs

Time (mins.) 60

Multiple TVs Same Content

Pan42A Pan42B

Query DB

vs

slide-37
SLIDE 37

Time (mins.)

15

= ?

Query DB

vs

TV

  • 1: 8

Movie

  • 1:20

Q.Len

  • 1:60
slide-38
SLIDE 38

Time (mins.)

15

= ?

Query DB

vs

TV

  • 1: 8

Movie

  • 1:20

Q.Len

  • 1:60
slide-39
SLIDE 39

Time (mins.)

15

= ?

Query DB

vs

TV

  • 1: 8

Movie

  • 1:20

Q.Len

  • 1:60

Best match

slide-40
SLIDE 40

Possible Query Results

  • Hit: the search engine is confident* in the match (accept)

and the match was the movie from which the query itself was extracted. (Success.)

  • Miss: the search engine is confident* in the match

(accept) but there is a mismatch between the search engine’s best guess and the query origin. (Failure.)

  • Reject: the best match was a not a clear winner*,

and the matching algorithm chooses not to respond. (Neither success nor failure.) * Confidence is the gap size between the top match and the runner up

slide-41
SLIDE 41

Query Length & Confidence

Reject

Confidence is the gap size between the top match and the runner up

Reject

Reject

Reject

Query Length

1 min

6 min 12 min 15 min

slide-42
SLIDE 42

Q4: Can we determine what is being watched from EMI?

slide-43
SLIDE 43

Q4: Can we determine what is being watched from EMI?

slide-44
SLIDE 44

Q5: Can we match lab EMI to home EMI?

slide-45
SLIDE 45

Home Details

Style Year Built Size Single family Home 2003 3000 sq. ft Apartment 2009 657 sq. ft. Multi-family Home 1906 800 sq. ft.

slide-46
SLIDE 46

Home Setup

slide-47
SLIDE 47

Home Setup

slide-48
SLIDE 48

Q5: Can we match lab EMI to EMI recorded in home settings

slide-49
SLIDE 49

Can we predict EMI from screen content without a TV?

Extensions:

slide-50
SLIDE 50

Predicting EMI

Input Features:

Extract Features Train Model Predict EMI

Model: Neural Net

  • Brightness: cumulative sum of averaged RGB intensities
  • Flux: change in brightness between consecutive frames
  • Edge Intensity: pixelsum of a Canny Edge filter
  • FFT: slope of the best fit line to an FFT
  • Color: mean and standard deviation for Gaussians fitted

to R, G, and B color histograms

  • Bitrate: kbits/second computed using FFMPEG

Cross Validation:

10 Train, 5 Test

slide-51
SLIDE 51

Predicting EMI

slide-52
SLIDE 52

Predicting EMI

slide-53
SLIDE 53

AC

@60Hz

use save TV EMI

SMPS

TV EMI: Theory of Operation

light scene dark scene

Backlight (brightness) is key source of TV EMI

slide-54
SLIDE 54

Defenses and Tensions

> Better filtering by power supplies > Signal Injection & Jamming (Energy Star) > Battery Masking [Mclaughlin 2011] > Infrequent sensor readings

slide-55
SLIDE 55

Takeaways

> Devices produce EMI on the powerline > Single sensor tracks many devices > EMI can be > Growing attack surface

slide-56
SLIDE 56

Questions?

miro@cs.washington.edu

slide-57
SLIDE 57

Does EMI matching scale?

slide-58
SLIDE 58

Does EMI matching scale?

slide-59
SLIDE 59

Sharp 32 and SNR

slide-60
SLIDE 60

“Our research shows that the analysis of the household’s electricity usage profile does reveal what channel the TV set in the household was displaying. “

Related Work

slide-61
SLIDE 61

Q1: repeatable EMI from repeated screen content?

Cross Correlation

A & B : Lion King C: Bourne Ultimatum

EMI Time (secs)

slide-62
SLIDE 62

58”

B Samsung 58-A

TVs (1)

slide-63
SLIDE 63

42”

Panasonic 42-A B

TVs (2)

slide-64
SLIDE 64

32”

B

Sharp 32-A

TVs (3)