Survey of Visible Light based Backscatter Communication...

34
Ju Wang Jan. 5, 2018 Survey of Visible Light based Backscatter Communication, Sensing 1

Transcript of Survey of Visible Light based Backscatter Communication...

Page 1: Survey of Visible Light based Backscatter Communication ...blizzard.cs.uwaterloo.ca/iss4e/wp-content/uploads/... · • Passive Sensing and Communication Using Visible Light: Taxonomy,

Ju Wang

Jan. 5, 2018

Survey of Visible Light based Backscatter Communication, Sensing

1

Page 2: Survey of Visible Light based Backscatter Communication ...blizzard.cs.uwaterloo.ca/iss4e/wp-content/uploads/... · • Passive Sensing and Communication Using Visible Light: Taxonomy,

Visible Light based Backscatter Communication

2

Page 3: Survey of Visible Light based Backscatter Communication ...blizzard.cs.uwaterloo.ca/iss4e/wp-content/uploads/... · • Passive Sensing and Communication Using Visible Light: Taxonomy,

Visible Light Backscatter Communication• Retro-VLC: Enabling Battery-free Duplex Visible Light Communication for Mobile and IoT

Applications, in HotMobile’15 (First Visible Light Backscatter Communication paper)

• PassiveVLC: Enabling Practical Visible Light Backscater Communication for Batery-free IoTApplications, in Mobicom’17 (An improvement paper for Retro-VLC)

• Pixelated VLC-Backscattering for Self-Charging Indoor IoT Devices, in IEEE PHOTONICS TECHNOLOGY LETTERS 2017 (An improvement paper for Retro-VLC)

• Passive Communication with Ambient Light, in CoNEXT’16 (A new idea for Visible Light Backscatter Communication)

• Battery-free Visible Light Sensing, in VLCS’17 (using visible light for sensing and RF Backscatter for communication)

• Passive Sensing and Communication Using Visible Light: Taxonomy, Challenges and Opportunities, in arXiv’17 (A short survey paper for Visible Light Backscatter Communication)

3

Page 4: Survey of Visible Light based Backscatter Communication ...blizzard.cs.uwaterloo.ca/iss4e/wp-content/uploads/... · • Passive Sensing and Communication Using Visible Light: Taxonomy,

Retro-VLC, in HotMobile’15

4

Page 5: Survey of Visible Light based Backscatter Communication ...blizzard.cs.uwaterloo.ca/iss4e/wp-content/uploads/... · • Passive Sensing and Communication Using Visible Light: Taxonomy,

• Motivation:

--Existing VLC systems focus on one-way communications, i.e., from illumination to mobile device.

--This paper enables a battery-free device to perform bi-directional communications over a shared light carrier.

• Basic idea:

--Retro-reflective fabric bounces light back to its source along its incoming direction.

5

Retro-VLC, in HotMobile’15

Page 6: Survey of Visible Light based Backscatter Communication ...blizzard.cs.uwaterloo.ca/iss4e/wp-content/uploads/... · • Passive Sensing and Communication Using Visible Light: Taxonomy,

• System architecture:

6

Retro-VLC, in HotMobile’15

Page 7: Survey of Visible Light based Backscatter Communication ...blizzard.cs.uwaterloo.ca/iss4e/wp-content/uploads/... · • Passive Sensing and Communication Using Visible Light: Taxonomy,

• System architecture:

7

Retro-VLC, in HotMobile’15

• Carrier: switching LED at 1MHz.

• Modulation: On/Off keying.

• Maximum distance: 2.6m.

• Field of view (FoV): ±25 degree.

• Uplink data rate: low, ~125 bps.

• Power consumption: reader, 12W; tag, 0.234mw.

Page 8: Survey of Visible Light based Backscatter Communication ...blizzard.cs.uwaterloo.ca/iss4e/wp-content/uploads/... · • Passive Sensing and Communication Using Visible Light: Taxonomy,

PassiveVLC, in Mobicom’17

8

“Seek to improve the data rate of RetroVLC ”

Page 9: Survey of Visible Light based Backscatter Communication ...blizzard.cs.uwaterloo.ca/iss4e/wp-content/uploads/... · • Passive Sensing and Communication Using Visible Light: Taxonomy,

PassiveVLC, in Mobicom’17

9

• Two observations:

--Manchester coding used in RetroVLC is not efciency.

--No need to “fully” switch on/off the LCD to convey a signal.

• PassiveVLC did two improvements:

--Replace Manchester coding with Miller code --> 8x data rate improvement.

Up to 1 kbps on a loss-free uplink within 1 m working distance

Page 10: Survey of Visible Light based Backscatter Communication ...blizzard.cs.uwaterloo.ca/iss4e/wp-content/uploads/... · • Passive Sensing and Communication Using Visible Light: Taxonomy,

PassiveVLC, in Mobicom’17

10

• Two observations:

--Manchester coding used in RetroVLC is not efciency.

--No need to “fully” switch on/off the LCD to convey a signal.

• PassiveVLC did two improvements:

--Design a trend-based modulation--> modulation time reducing, 4 msto 1 ms.

Page 11: Survey of Visible Light based Backscatter Communication ...blizzard.cs.uwaterloo.ca/iss4e/wp-content/uploads/... · • Passive Sensing and Communication Using Visible Light: Taxonomy,

PassiveVLC, in Mobicom’17

11

• I think the technical contributions of this paper are weak, they did not solve the basic challenges:

--Can not work in NLOS, long range, large beam angle.

--Easily distorted by other light sources.

--Still low data rate.

Page 12: Survey of Visible Light based Backscatter Communication ...blizzard.cs.uwaterloo.ca/iss4e/wp-content/uploads/... · • Passive Sensing and Communication Using Visible Light: Taxonomy,

Pixelated VLC-Backscattering in IEEE PHOTONICS TECHNOLOGY LETTERS 2017

12

“Seek to improve the data rate of RetroVLC ”

Page 13: Survey of Visible Light based Backscatter Communication ...blizzard.cs.uwaterloo.ca/iss4e/wp-content/uploads/... · • Passive Sensing and Communication Using Visible Light: Taxonomy,

Pixelated VLC-Backscattering in IEEE PHOTONICS TECHNOLOGY LETTERS 2017

13

(b) The proposed pixelated based system.(a) Past VLC backscatter system architecture.

Multiple smaller VLC backscatters are integrated to generate multi-level signals

Page 14: Survey of Visible Light based Backscatter Communication ...blizzard.cs.uwaterloo.ca/iss4e/wp-content/uploads/... · • Passive Sensing and Communication Using Visible Light: Taxonomy,

Ambient Light Passive Communication, in CoNEXT’16

14

Page 15: Survey of Visible Light based Backscatter Communication ...blizzard.cs.uwaterloo.ca/iss4e/wp-content/uploads/... · • Passive Sensing and Communication Using Visible Light: Taxonomy,

Ambient Light Passive Communication, in CoNEXT’16

15

• Key idea:

--Embed data into reflective objects (with different reflective materials) for passive communication

--When object moves, intensity of reflected light changes, altering the light impinging towards receiver.

Page 16: Survey of Visible Light based Backscatter Communication ...blizzard.cs.uwaterloo.ca/iss4e/wp-content/uploads/... · • Passive Sensing and Communication Using Visible Light: Taxonomy,

Battery-free Visible Light Sensing, in VLCS’17

16

• Using solar cells as light sensors. Ambient light as a sensing medium allows tracking of shadows cast by objects.

• Working distance & power: 330m, at a peak power of 20 μWs.

Page 17: Survey of Visible Light based Backscatter Communication ...blizzard.cs.uwaterloo.ca/iss4e/wp-content/uploads/... · • Passive Sensing and Communication Using Visible Light: Taxonomy,

Visible Light based Sensing

17

Page 18: Survey of Visible Light based Backscatter Communication ...blizzard.cs.uwaterloo.ca/iss4e/wp-content/uploads/... · • Passive Sensing and Communication Using Visible Light: Taxonomy,

Key Groups

18

Microsoft Research Asia

Feng Zhao Xinyu Zhang

University of California San Diego

Xia Zhou Tian He

Dartmouth College University of Minnesota

• Focuses on Visible Light based localization

• Proposed First Visible Light Backscatter Communication

• Proposed First Visible Light localization

• Focuses on both Visible Light based gesture recognition andCommunication

• Few papers about Visible Light based localization

Page 19: Survey of Visible Light based Backscatter Communication ...blizzard.cs.uwaterloo.ca/iss4e/wp-content/uploads/... · • Passive Sensing and Communication Using Visible Light: Taxonomy,

19

The First Visible Light based localization paperNSDI’14

Microsoft Research AsiaFeng Zhao

Page 20: Survey of Visible Light based Backscatter Communication ...blizzard.cs.uwaterloo.ca/iss4e/wp-content/uploads/... · • Passive Sensing and Communication Using Visible Light: Taxonomy,

Epsilon: A Visible Light Based Positioning System, in NSDI’14

System setup and Performance

• Method: Range-based localization.

• TX: 5 Commercial LEDs.

• RX: 1 Smart phone with light sensor.

• Test area: Office, ~ 3m x 3.5m.

• Accuracy: ~ 0.7m

20

Page 21: Survey of Visible Light based Backscatter Communication ...blizzard.cs.uwaterloo.ca/iss4e/wp-content/uploads/... · • Passive Sensing and Communication Using Visible Light: Taxonomy,

Epsilon: A Visible Light Based Positioning System, in NSDI’14

Key Contribution:

• An Optical Channel Model that shows the relationship between light RSS and distance

21

Weakness:

• The model only works well within ± 60°

• Requiring light sensor faces squarely upward toward ceiling

Θ=φ=0

Page 22: Survey of Visible Light based Backscatter Communication ...blizzard.cs.uwaterloo.ca/iss4e/wp-content/uploads/... · • Passive Sensing and Communication Using Visible Light: Taxonomy,

MobiCom’14 The First AoA based Visible Light localization paper

Page 23: Survey of Visible Light based Backscatter Communication ...blizzard.cs.uwaterloo.ca/iss4e/wp-content/uploads/... · • Passive Sensing and Communication Using Visible Light: Taxonomy,

Luxapose, in MobiCom’14

23

System setup and Performance• Method: AoA-based localization.• TX: 5 Commercial LEDs with known locations as landmarks.• RX: 1 Smart phone with camera.• Test area: office room, ~ 71.1 cm x 73.7 cm.• Accuracy: ~ 7 cm

Page 24: Survey of Visible Light based Backscatter Communication ...blizzard.cs.uwaterloo.ca/iss4e/wp-content/uploads/... · • Passive Sensing and Communication Using Visible Light: Taxonomy,

MobiCom’15 A high localization accuracy paper with errors ~7 mm

Xinyu Zhang

Page 25: Survey of Visible Light based Backscatter Communication ...blizzard.cs.uwaterloo.ca/iss4e/wp-content/uploads/... · • Passive Sensing and Communication Using Visible Light: Taxonomy,

Near-Field Visible Light Sensing, in Mobicom’15

25

System setup and Performance• Method: RSS-model based Location.• TX: 1 LED with known location.• RX: 2 photodetectors (PDs) with known locations.• Test area: office room, ~ 8 cm x 8 cm.• Accuracy: ~ 7 mm

Page 26: Survey of Visible Light based Backscatter Communication ...blizzard.cs.uwaterloo.ca/iss4e/wp-content/uploads/... · • Passive Sensing and Communication Using Visible Light: Taxonomy,

MobiCom’17

Xinyu Zhang

The AoA based Visible Light localization without a camera

Page 27: Survey of Visible Light based Backscatter Communication ...blizzard.cs.uwaterloo.ca/iss4e/wp-content/uploads/... · • Passive Sensing and Communication Using Visible Light: Taxonomy,

Pulsar, in Mobicom’17

27

System setup and Performance• Method: AoA-based localization.• TX: 4 fluorescent/LED light with known locations.• RX: 2-PD based light sensor.• Test area: indoor, ~ ? cm x ? cm.• Accuracy: ~ 20 cm

Motivation:

Camera is expensive;

PD is much cheaper.

Self-built receiver

Page 28: Survey of Visible Light based Backscatter Communication ...blizzard.cs.uwaterloo.ca/iss4e/wp-content/uploads/... · • Passive Sensing and Communication Using Visible Light: Taxonomy,

Key Contribution:

• Using RSS to calculate AoA

28

Pulsar, in Mobicom’17

The Ratio of RSS at two near by PDs

AoA

Page 29: Survey of Visible Light based Backscatter Communication ...blizzard.cs.uwaterloo.ca/iss4e/wp-content/uploads/... · • Passive Sensing and Communication Using Visible Light: Taxonomy,

Sensys’17

Tian He

Page 30: Survey of Visible Light based Backscatter Communication ...blizzard.cs.uwaterloo.ca/iss4e/wp-content/uploads/... · • Passive Sensing and Communication Using Visible Light: Taxonomy,

SmartLight, in Sensys’17

30

System setup and Performance• Method: geometrical optical property based localization.• TX: 1 modified LED array with known location.• RX: 1 light sensor.• Test area: office room, ~ 2.5 m x 2.5 m.• Accuracy: ~ 10 cm

Lens

Basic idea:

• Each LED pixel blinks at a unique frequency.

• Locations of LEDs are known for light sensor.

• A sensor receives light from a part of LEDs.

• The location can be estimated based on the geometrical optical property.

Page 31: Survey of Visible Light based Backscatter Communication ...blizzard.cs.uwaterloo.ca/iss4e/wp-content/uploads/... · • Passive Sensing and Communication Using Visible Light: Taxonomy,

MobiCom’15

Xia Zhou

Page 32: Survey of Visible Light based Backscatter Communication ...blizzard.cs.uwaterloo.ca/iss4e/wp-content/uploads/... · • Passive Sensing and Communication Using Visible Light: Taxonomy,

Human Sensing Using VLC, in Mobicom’15

32

System setup and Performance• Method: Mapping shadow to gesture.• TX: 5 LED with known location.• RX: 324 photodetectors (PDs) with known locations.• Test area: office room, ~ 3m x 3m.• Accuracy: more than 90%

Page 33: Survey of Visible Light based Backscatter Communication ...blizzard.cs.uwaterloo.ca/iss4e/wp-content/uploads/... · • Passive Sensing and Communication Using Visible Light: Taxonomy,

Human Sensing Using VLC, in Mobicom’15

33

Challenge 1: Separating shadows caused by multiple LEDs.Solution: Using the frequency differences of different LEDs.

Solution: Using skeleton model and optimization.

Challenge 2: Mapping 2D shadows to 3D gesture.

Page 34: Survey of Visible Light based Backscatter Communication ...blizzard.cs.uwaterloo.ca/iss4e/wp-content/uploads/... · • Passive Sensing and Communication Using Visible Light: Taxonomy,

Summary

34

• High accuracy: easily achieving sub-meter accuracy due to LoS propagations.

• Limited coverage

--Cost of LED, i.e., requiring unique frequency.

--Limited sensing ability of AoA/Range, e.g., only works within ±60°

High accuracy, but only works within a limited range.