ASLP: An Adaptive Framework for Spatial Location Privacy

This research investigates Adaptive Spatial Location Privacy (ASLP), a new approach to harden location privacy while preserving the utility of data. Unlike basic vicinity-based methods, ASLP employs dynamic tuning and countermeasures to provide a superior privacy-utility trade-off across varying risk levels. The novelty of this work lies in the adaptive nature of ASLP, which adjusts privacy levels based on context, and the incorporation of countermeasures like Non-Uniform Transform and Asymmetric Noise to mitigate potential attacks. Experiments demonstrate that ASLP significantly outperforms baseline VICINITY-PROTECTION in terms of Location Indistinguishability (reaching 0.6-0.65 at medium risk), Anonymity Set Size (0.1-0.13) and Distance to Nearest Neighbor Privacy (DNNP) (0.6-0.65), while maintaining high utility (85- 95% percentage within threshold 0.02).Countermeasures, especially the Non-Uniform Transform, further enhance privacy, although they may come at a slight cost to utility.