Privacy-Preserving Distributed Edge Computing with Guardians

Project Title:

Privacy-Preserving Distributed Edge Computing with Guardians

Supervisor(s):

Iqbal Gondal, Alireza Jolfaei,

Contact person and email address:

Iqbal Gondal, Iqbal.gondal@federation.edu.au, Director ICSL

A brief description of the project:

The purpose of this project is to explore novel mechanisms to protect the privacy of data providers in the edge of the network, in the face of misconfigurations, conflicts, adversarial attacks, and evolution of the IoT environment itself. IoT devices input a continuous stream of sensed information (e.g., health record and real-time physiological signals) to the Edge server (e.g., a patient agent who runs data analytics on behalf of the patient). The edge processing algorithm (which could be a trained ML/AI model, or an inference model or an estimation algorithm) derives logical conclusions or makes predictions w.r.t. the sensed information with a certain precision. The edge server sends its processing outcomes to the cloud (e.g., Healthcare) for further processing.

In this project, we will develop a Guardian which monitors the input and output of Edge servers and enforces dynamic privacy settings for data providers to maintain an optimal privacy value obtained by the privacy-utility trade-off. We will further investigate resilience of our proposed model in the presence of APT s.