Dejavu Outdoor localization system Provides accurate and energy-efficient outdoor localization Uses only energy-efficient sensors or sensors which are already running leverages road landmarks like moving over potholes, bumps, tunnels etc
Introduction GPS considered as de facto standard for outdoor localization Dejavu's approach is based on low-energy sensors (accelerometer,gyroscope,compass) Using array of sensors it identifies landmarks (anchors) crowdsourcing to build database
System overview
Raw sensor information System collects raw sensor information cellular network information (RSS and associated cell tower ID) opprtunistically leverags the Wifi chip to collect surrounding Wifi APs
Error Resetting uses Vincenty's formula to limit accumulated error system uses physical and virtual anchors
Physical and virtual anchors Dejavu uses two types of anchors physical anchors mapped to road features like bridges, tunnels, speed-bumps extracted from map or through pror knowledge includes points with unique GSM or Wifi RSS learned through crowd-sourcing
Anchor detection large number of road features can be identified on their unique signature
Physical anchors used to seed the anchor database identified from maps or by prior knowledge of the location
Physical anchors Curves and turns - - road curvature forces car to change its direction which results in big variance of phone's orientation
Tunnels - - drop in cellular signal
- large variance in the ambient magnetic field in x-direction
Bridges - - cars go up at the start of the bridge and go down at the end
- reflected in x-gravity or y-gravity acceleration
- cat's eye does not cause high variance in y or z-axis gravity accleration
- speed bumps usually have highest variance
- railway crossing leads to medium variance
Virtual anchors uses un-supervised learning techniques to identify virtual anchors anomaly detection techniques are used to identify anomalies in sensor readings. they are clustered into sensor space to identify candidate clusters points of each cluster are spatially clustered to identify the location of each anchor
Anomaly detection h - bandwidth n - sample size K - kernel function
Two stage clustering Cluster feature space using hierarchical clustering in vector feature space spatial clustering of points in clusters obtained previously cluster is accepted if the number of points is above a threshold
Computing anchor location
Discussion Anchors aliasing - classes of anchors can be confused with other anchors
- leverages map context information
- limits search space to small area around the user location
Processing location - can be split into a client-server architecture
Other sensors - other sensors of the phone such as camera, mic could be used
Handling heterogeneity
Evaluation
Anchor localization accuracy
Effect of anchor density on accuracy
Dejavu against other systems
Dostları ilə paylaş: |