PrivacyPreserving Trajectory Collection

1
Privacy-PreservingTrajectory Collection

• Gyozo Gidófalvi
• Uppsala University, Dept. of Information
  Technology
• Geomatic ApS
• Xuegang Huang and Torben Bach Pedersen
• Aalborg University, Dept. of Computer Science

2
Problem Setting

• Premise Accurate trajectory patterns necessary
  for LBSs
• Task Collect exact trajectories of mobile users
  in a privacy-preserving manner
• Objective Use free, energy-saving, short-range
  wireless P2P communication (Bluetooth, ZigBee)
• Problem Hardware ID is exposed in P2P
  communication

Need to break the link between public and private
information!
Data item
HID
Private trajectory
3
Location Privacy Definitions
DEF. k-a-anonymity k-anonymity a-diversity
4
Privacy-Preserving Trajectory Collection in Five
Stages
Server
Merge traj pieces and monitor of copies
After Ts2t if the majority parity is
odd?discard even?store
1 2 2
5
Results

• Realistic simulation shows that the method works
  under reasonable conditions and anonymity
  settings
• gt 98 of the clients report their collected data
• gt 95 of the data is reported
• Not reported data is recent and can be corrected
  for
• Method / system that
• collects exact trajectories without a loss
• does not require trusted components
• and provides strong privacy guarantees