Abstract:Federated learning based participatory sensing has gained much attention lately for the vital task of urban sensing due to privacy and security issues in conventional machine learning. However, inference attacks by the honest-but-curious application server or a malicious adversary can leak the personal attributes of the participants, such as their home and workplace locations, routines, and habits. Approaches proposed in the literature to prevent such information leakage, such as secure multi-party computation and homomorphic encryption, are infeasible for urban sensing applications owing to high communication and computation costs due to multiple rounds of communication between the user and the server. Moreover, for effective modeling of urban sensing phenomenon, the application model needs to be updated frequently - every few minutes or hours, resulting in periodic data-intensive updates by the participants, which severely strains the already limited resources of their mobile devices. This paper proposes a novel low-cost privacy-preserving framework for enhanced protection against the inference of participants' personal and private attributes from the data leaked through inference attacks. We propose a novel approach of strategically leaking selected location traces by providing computation and communication-light direct (local) model updates, whereas the rest of the model updates (when the user is at sensitive locations) are provided using secure multi-party computation. We propose two new methods based on spatiotemporal entropy and Kullback–Leibler divergence for automatically deciding which model updates need to be sent through secure multi-party computation and which can be sent directly. The proposed approach significantly reduces the computation and communication overhead for participants compared to the fully secure multi-party computation protocols. It provides enhanced protection against the deduction of personal attributes from inferred location traces compared to the direct model updates by confusing the application server or malicious adversary while inferring personal attributes from location traces. Numerical experiments on the popular Geolife GPS trajectories dataset validate our proposed approach by reducing the computation and communication requirements by the participants significantly and, at the same time, enhancing privacy by decreasing the number of inferred sensitive and private locations of participants.

Privacy-preserving computation of participatory noise maps in the cloud

Participatory Privacy: Enabling Privacy in Participatory Sensing

Privacy-Aware High-Quality Map Generation with Participatory Sensing

Socially Privacy-Preserving Data Collection for Crowdsensing

Crowd-Empowered Privacy-Preserving Data Aggregation for Mobile Crowdsensing.

Incentivizing Crowdsensing-Based Noise Monitoring with Differentially-Private Locations

Privacy-Preserving Data Aggregation for Mobile Crowdsensing with Externality: an Auction Approach

Extended Capabilities for a Privacy-Enhanced Participatory Sensing Infrastructure (PEPSI)

Privacy Preserving RSS Map Generation for a Crowdsensing Network

A New Noise Generating Method Based on Gaussian Sampling for Privacy Preservation

Privacy in Cloud Computing through Immersion-based Coding

Efficient and Privacy-Preserving Non-Interactive Truth Discovery for Mobile Crowdsensing.

PAS-MC: Privacy-preserving Analytics Stream for the Mobile Cloud

Cloud-Enabled Privacy-Preserving Truth Discovery in Crowd Sensing Systems

Computation and communication efficient approach for federated learning based urban sensing applications against inference attacks

Privacy-preserving Computation for Location-Based Information Survey Via Mobile Cloud Computing

Privacy-Preserving Truth Discovery in Crowd Sensing Systems

Differentially Private Set-Based Estimation Using Zonotopes

Privacy Preserving Semi-Decentralized Mean Estimation over Intermittently-Connected Networks

Privacy-Preserving Distributed Analytics in Fog-Enabled IoT Systems

Proximity-Aware Local-Recoding Anonymization with MapReduce for Scalable Big Data Privacy Preservation in Cloud