DATAMAN MOBILE COMPUTING LABORATORY VOR Base Stations for Indoor - PowerPoint PPT Presentation

DATAMAN MOBILE COMPUTING LABORATORY VOR Base Stations for Indoor 802.11 Positioning Dragos ¸ Niculescu and Badri Nath { dnicules,badri } @cs.rutgers.edu

☞ indoor positioning existing systems require either: ❍ extra infrastructure + good accuracy - instrumentation - specialized beacons, badges - LOS ❍ signal strength (SS) map + existing 802.11 base stations - map depends on people, furniture, ... - centralized database Dragos ¸ Niculescu – VOR Base Stations for Indoor 802.11 Positioning

signal strength map (example) ☞ 25m (0,0) 56m = basestations = sample point RADAR project (Microsoft) 1. build SS map: ❍ for each point, measure SS to all 5 BS 2. query: ❍ measure SS to 5 BS ⇒ best match in the map Dragos ¸ Niculescu – VOR Base Stations for Indoor 802.11 Positioning

☞ VORBA - VOR BAse stations goals: ❍ no signal strength map ❍ less infrastructure ❍ move complexity to the 802.11 base station ❍ use: − angles − ranges − angles and ranges Dragos ¸ Niculescu – VOR Base Stations for Indoor 802.11 Positioning

☞ VORBA prototype 90° 120° 60° IR receiver IR sender IR sender 150° 30° antenna 0° − −4 4 0− 0− 3 3 0 0 − − 2 2 0 0 − − 1 1 0 0 − − 3 0 3 0 180° 802.11 card 802.11 card 330° 210° 300° 240° 270° prototype base station directional antenna pattern Dragos ¸ Niculescu – VOR Base Stations for Indoor 802.11 Positioning

☞ basic idea −62 π −64 2 signal strength [dBm] −66 −68 mean SS π 2 π −70 −72 −74 3 π 2 −76 −78 π 3 π 2 π π 2 2 signal strength variation = SS ( α ) � angle � peak → angle 1 . SS ( α ) 2 . 3 . and / or → position mean → range range Dragos ¸ Niculescu – VOR Base Stations for Indoor 802.11 Positioning

☞ experiments ❍ 32 measurement points ❍ 5 + 2 base stations ❍ N/E/S/W measurements of 3-4 revolutions each Dragos ¸ Niculescu – VOR Base Stations for Indoor 802.11 Positioning

☞ angles only positioning 1 cumulative probability 0.75 0.5 0.25 using best angle using first two angles 0 0 2 4 6 8 10 12 14 16 error in meters ❍ 3.5m median position error ❍ 3m if we knew the best peak Dragos ¸ Niculescu – VOR Base Stations for Indoor 802.11 Positioning

☞ quantized angles −500 0 500 1000 1500 2000 2500 −500 0 500 1000 1500 2000 2500 1500 1500 1400 1400 1200 1200 1000 1000 1000 1000 552x908 + o 552x908 o 1650x794 + + + + 800 800 + + 1650x794 + + 600 600 500 500 400 400 814x358 814x358 200 200 218x178 218x178 + 1776x130 1776x130 0 0 0 0 −200 −200 −400 −400 −500 −500 −400 −200 0 200 400 600 800 1000 1200 1400 1600 1800 2000 2200 2400 −400 −200 0 200 400 600 800 1000 1200 1400 1600 1800 2000 2200 2400 ❍ measurements rounded to the nearest 45 ◦ ❍ simulation ❍ little degradation for 45 ◦ and 22 . 5 ◦ quantizations Dragos ¸ Niculescu – VOR Base Stations for Indoor 802.11 Positioning

☞ angles & ranges σ r σ y σ a σ x A M r ❍ angle error σ a = 0 . 4 radians ≃ 21 ◦ ❍ range error σ r = 0 . 2 r ❍ approximate uncertainty as an ellipse ❍ error ellipse increases with distance Dragos ¸ Niculescu – VOR Base Stations for Indoor 802.11 Positioning

☞ angles & ranges uncertainty how to combine several readings? Kalman filter . BS 1 α 1 ρ 1 BS 3 BS 2 Dragos ¸ Niculescu – VOR Base Stations for Indoor 802.11 Positioning

☞ angles & ranges positioning 1 cumulative probability 0.75 0.5 0.25 7 BS 5 BS 3 BS 1 BS 0 0 16 4 8 10 12 2 6 14 error in meters ❍ more base stations ⇒ better positions ❍ 2.1m median position error (all 7 BS) Dragos ¸ Niculescu – VOR Base Stations for Indoor 802.11 Positioning

☞ summary ❍ VORBA = VOR base station ❍ complexity into the base station − less infrastructure − no SS map ❍ revolving basestation measures SS ( α ) to derive − discrete angles − angle distributions − ranges ❍ works with quantized angles as well ❍ can achieve 2.1m - 4m median error Dragos ¸ Niculescu – VOR Base Stations for Indoor 802.11 Positioning

☞ index ❍ indoor positioning ❍ angles only positioning − angulation/lateration − discrete angles − SS map example − quantized angles ❍ VOR BAse station ❍ angles and ranges − prototype − uncertainty − basic idea − performance − experiment setup ❍ summary Dragos ¸ Niculescu – VOR Base Stations for Indoor 802.11 Positioning

☞ trilateration N N N β B A γ α C M ( x M − x A ) 2 + ( y M − y A ) 2 = MA 2 ( x M − x B ) 2 + ( y M − y B ) 2 = MB 2 solve for ( x M , y M ) ( x M − x C ) 2 + ( y M − y C ) 2 = MC 2 ❍ MA, MB, MC are affected by errors ❍ several methods available Dragos ¸ Niculescu – VOR Base Stations for Indoor 802.11 Positioning

☞ triangulation N N N β B A γ α C M ( x M − x A ) sin α = ( y M − y A ) cos α solve for ( x M , y M ) ( x M − x B ) sin β = ( y M − y B ) cos β ( x M − x C ) sin γ = ( y M − y C ) cos γ ❍ α, β, γ -affected by errors (Gaussian) ❍ several methods available Dragos ¸ Niculescu – VOR Base Stations for Indoor 802.11 Positioning

☞ ranges and angles N N N β B A γ α C M x M = x A + MA cos α = x B + MB cos β = x C + MC cos γ y M = y A + MA sin α = y B + MB sin β = y C + MC sin γ ❍ one base station is theoretically enough ❍ α, β, γ, MA, MB, MC - affected by errors Dragos ¸ Niculescu – VOR Base Stations for Indoor 802.11 Positioning

☞ best peak distribution 0.6 0.6 histogram of number of peaks Histogram of SS rank mean = 4.5 peaks of best peak probability 60% 30% 0 1 2 3 4 5 6 7 8 −1 0 1 2 3 4 5 6 7 8 peak rank number of peaks ❍ 4.5 peaks on average ❍ best peak is first/second 90% of the time Dragos ¸ Niculescu – VOR Base Stations for Indoor 802.11 Positioning

☞ other peak distribution true direction true direction π − π 4 4 15% 33% − π π π − π − π π 2 2 4 4 ❍ other peaks point away from true direction Dragos ¸ Niculescu – VOR Base Stations for Indoor 802.11 Positioning

☞ triangulation analysis 0.02 0.02 0.002 Var[x] Var[x] Var[x] 0.01 0.001 0.01 simulation simulation simulation lower bound lower bound lower bound 0 0 0 0 0.6 1.2 1.8 2.4 3 3.6 4.2 4.8 0.1 0.15 0.2 0.25 0.3 0.35 0.4 0.45 0.5 0.55 0.6 0.65 0 0.05 0.1 0.15 0.2 0.25 0.3 0.35 0.4 0.45 0.5 1 σ 2 1 a R λ ln Rm σ 2 a V ar [ x ] > λπ ln R R m ❍ V ar [ x ] - standard dev. of positioning error ❍ λ - density of basestations / m 2 ❍ to improve positioning: 1. decrease measurement error σ a 2. use more basestations Dragos ¸ Niculescu – VOR Base Stations for Indoor 802.11 Positioning

☞ angle distribution Dragos ¸ Niculescu – VOR Base Stations for Indoor 802.11 Positioning

☞ quantized angles 1 cumulative probability 0.75 0.5 0.25 best angle (non quantized) quantization 45 quantization 22.5 quantization 90 0 14 0 2 4 6 8 10 12 error in meters ❍ little degradation for − 16 directions ( 22 . 5 ◦ ) − 8 directions ( 45 ◦ ) Dragos ¸ Niculescu – VOR Base Stations for Indoor 802.11 Positioning

☞ range inference ❍ open space attenuation: SS [ dBm ] = SS 0 [ dBm ] − log 10 ( d d 0 ) n ❍ d ( SS ) − obtained through fitting − known to be unreliable ❍ we obtain it from integration of SS ( α ) ❍ 5-fold cross validation − corridor basestations - waveguide effect − median range error 2.8m Dragos ¸ Niculescu – VOR Base Stations for Indoor 802.11 Positioning

☞ positioning w. ranges 1 cumulative probability 0.75 0.5 0.25 0 14 0 2 8 10 12 4 6 error in meters ❍ trilateration 5 base stations ❍ median position error 4.5m Dragos ¸ Niculescu – VOR Base Stations for Indoor 802.11 Positioning

☞ discussion ❍ triangulation with large outliers ❍ use more than two angles? ❍ no correlation between − angle error and distance − angle error and SS ❍ corridors ⇒ waveguides ❍ revolving signal at the mobile? ❍ data performance? Dragos ¸ Niculescu – VOR Base Stations for Indoor 802.11 Positioning