Datasets
Standard Dataset
SCDC
- Citation Author(s):
- Submitted by:
- Qi Deng
- Last updated:
- Mon, 01/27/2025 - 05:17
- DOI:
- 10.21227/8psc-qa22
- License:
12 Views- Categories:
- Keywords:
Abstract
Abstract—Sparse Mobile CrowdSensing is an efficient data collection paradigm that recruits participants to gather data from partial spatiotemporal regions and leverages inherent correlations among these data to infer the remaining uncollected data. However, enabling accurate inference requires participants to upload sensitive spatiotemporal information, which poses significant privacy leakage risks. Traditional methods address these risks by obfuscating the uploaded spatial data, but this often compromises inference accuracy. To tackle this challenge, we propose a novel spatiotemporal obfuscation strategy that obfuscates collected data to more critical spatiotemporal regions and adjusts the values based on spatiotemporal correlations, ensuring high inference accuracy under privacy constraints. Specifically, we employ an active learning-based exponential mechanism for spatial obfuscation. After spatial obfuscation, an association model is applied to capture spatiotemporal relationships and adjust data values. Then we use a Fourier transform-based switch operation for temporal obfuscation, both designed to guide data towards regions most beneficial for inference, to ensure high-quality completion while maintaining privacy constraints. Experimental evaluations demonstrate that our strategy significantly enhances data completion quality in urban sensing tasks, such as environmental monitoring and traffic management, while providing robust privacy protections.
L-SRR: L-SRR perturbs input positions by employing a stepwise probability distribution that adapts the perturbation probability based on the spatial distance between the input and output positions. In contrast to traditional LDP mechanisms, L-SRR assigns higher probabilities to output positions that are closer to the input location, thereby enhancing the utility of the obfuscated data. The algorithm organizes output positions into groups based on distance, and the perturbation probabilities follow a stepwise distribution across these groups. This distanceaware perturbation strategy ensures that the obfuscated data retains a higher level of spatial relevance while still preserving privacy [29]. • LPDO: LPDO adopts a two-step process to balance data privacy and utility. First, it learns a data adjustment function that aligns the raw sensor data with perturbed positions. Next, a linear programming framework is employed to select the optimal position obfuscation function, which minimizes uncertainty in the data adjustment process. To further enhance data completion accuracy, LPDO incorporates an uncertainty-aware inference algorithm that leverages the perturbed data for improved reconstruction. This focus on optimal obfuscation and uncertainty-aware inference makes LPDO a robust approach for privacypreserving spatial data processing [22]. • HMPC: HMPC utilizes the Hilbert mapping technique to transform two-dimensional spatial coordinates into onedimensional Hilbert indices. By perturbing these indices instead of directly obfuscating the spatial coordinates, HMPC introduces less Laplace noise, thereby preserving more spatial data utility. A notable feature of the Hilbert mapping approach is its sensitivity to small changes in indices, which can result in large variations in spatial distances. This allows HMPC to achieve a balance between privacy protection and utility, as smaller perturbations can still provide significant spatial obfuscation [30].
