Wireless Mesh Networks for Localization Dr Kristofer Pister

  • Slides: 16
Download presentation
Wireless Mesh Networks for Localization Dr. Kristofer Pister Prof. EECS, UC Berkeley Founder &

Wireless Mesh Networks for Localization Dr. Kristofer Pister Prof. EECS, UC Berkeley Founder & CTO, Dust Networks

Outline Goals Challenges Technical Approaches Real-world Deployments Future Technology Developments 1

Outline Goals Challenges Technical Approaches Real-world Deployments Future Technology Developments 1

First RFID myth “Now I’ll know where all of my stuff is, all of

First RFID myth “Now I’ll know where all of my stuff is, all of the time!” 2

First RFID myth “Now I’ll know where all of my stuff is, all of

First RFID myth “Now I’ll know where all of my stuff is, all of the time!” The reality: - RFID systems tell you where your stuff was the last time it went by a reader. Very useful if: - You have chokepoints in your supply chain - You can put a reader on ALL possible entries/exits - You can serialize your material flow 3

Second RFID myth “Now I’ll know where all of my stuff is, all of

Second RFID myth “Now I’ll know where all of my stuff is, all of the time!” “And it will only cost ten cents per tag!” 4

Second RFID myth “Now I’ll know where all of my stuff is, all of

Second RFID myth “Now I’ll know where all of my stuff is, all of the time!” “And it will only cost ten cents per tag!” The Reality: - Tags may get to $0. 10 or lower (probably $0. 25 today) and… - Readers may get to $100 but… - Installing power and comm to all of the readers will dominate the cost of tags for most deployments 5

Mesh to the rescue? Good news, bad news + Low cost of “reader” installation

Mesh to the rescue? Good news, bad news + Low cost of “reader” installation - Completely wireless - Self-configuring - 5 -10 year battery life for infrastructure + Sense state of items as well as location - Increased tag cost - Dollars, not pennies - Constraints on measurement accuracy 6

RSSI-based localization Received Signal Strength Indication - Used as a surrogate for range Use

RSSI-based localization Received Signal Strength Indication - Used as a surrogate for range Use range to known locations to estimate position - Tri-lateration or - Multi-lateration - (not Triangulation) 7

Other localization approaches Acoustic - Sub-cm accuracy in unobstructed environments GPS - Need clear

Other localization approaches Acoustic - Sub-cm accuracy in unobstructed environments GPS - Need clear line of sight to satellites Pseudolites - Still doesn’t work well indoors - Rosum? 8

Distance vs. Received Signal Strength RSSI and distance for Consolidated network 1/R 2? 1/R

Distance vs. Received Signal Strength RSSI and distance for Consolidated network 1/R 2? 1/R 4? Distance [meters] 60 40 20 0 -100 -90 -80 -70 -60 RSSI [d. Bm] -50 -40 9

Distance vs. Received Signal Strength RSSI and distance for Consolidated network 1/R 2? 1/R

Distance vs. Received Signal Strength RSSI and distance for Consolidated network 1/R 2? 1/R 4? Distance [meters] 60 40 20 0 -100 -90 -80 -70 -60 RSSI [d. Bm] -50 -40 10

Distance vs. Received Signal Strength RSSI and distance for Consolidated network 1/R 2? 1/R

Distance vs. Received Signal Strength RSSI and distance for Consolidated network 1/R 2? 1/R 4? Distance [meters] 60 40 20 0 -100 -90 -80 -70 -60 RSSI [d. Bm] -50 -40 11

Be cautious of RSSI Chipcon CC 2341 802. 15. 4 datasheet: - 0. 5

Be cautious of RSSI Chipcon CC 2341 802. 15. 4 datasheet: - 0. 5 m resolution - “Location error can be less than 3 m, depending on factors described below…” - Requires calibrated base stations 10 m accuracy reasonable, 1 m no way Maybe it’s perfect for you, but… - Get a demonstration in your environment - Run it for a while and move stuff around 12

Scalability: 32 acre Test Network 1, 100 m -1400 Motes - 32 Acres -

Scalability: 32 acre Test Network 1, 100 m -1400 Motes - 32 Acres - ~3 m accuracy? 600 m 13

Technology to the rescue Accurate RF distance measurement UWB - 10 cm? RF Round-Trip

Technology to the rescue Accurate RF distance measurement UWB - 10 cm? RF Round-Trip Time of Flight - < 1 m 14

Outdoor Measurement Results Steven Lanzisera, UC Berkeley - slanzise@eecs. berkeley. edu - pister@eecs. berkeley.

Outdoor Measurement Results Steven Lanzisera, UC Berkeley - slanzise@eecs. berkeley. edu - pister@eecs. berkeley. edu 15