Nearby friend alert: Location anonymity in mobile geosocial networks

Mobile geosocial networking services could be the killer app for next-generation mobile computing. However, the privacy issue - in particular, location privacy - has both users and government authors concerned. The authors address this issue for the 'nearby friend alert' service, common in mobile geosocial networks. They review representative works on privacy-preserving proximity detection and present a new quantitative solution. They adopt the grid-and-hashing paradigm and develop optimal grid overlay and multilevel grids to increase the detection accuracy while saving the wireless bandwidth. Based on these techniques, they devise the client-side location update scheme and the server-side update handling procedure for continuous proximity detection. Simulation results show that their approach is efficient and scalable under various system parameters and user moving speeds.