Design of Advanced Privacy Preserving Model for Protecting Privacy within a Fog Computing Scenario
Pinkal Jain, Vikas Thada, Ajay Lala · 2023
Due to the ongoing development of computer and communication abilities, the IoT is becoming more and more significant in many smart applications. As a result, IoT devices produce plenty of data every day, which provides a strong basis for ML to succeed. The IoT data's strict privacy standards, however, make its machine learning extremely challenging. Numerous privacy-based ML strategies have been developed to safeguard the privacy of data. The majority of current schemes do not provide generic answers and only focus on specific models, which is not the best option for engineering practice. To solve this issue, we proposed a powerful machine learning model that protects privacy within a fog computing scenario. The software service provider (SSP) can train models while protecting the privacy of the data on the fog nodes using the APPML (Advanced Privacy Preserving Machine Learning) framework. Only SSP may access the model parameters, and it is possible to maintain the privacy of data stored at the fog nodes. Experimental results demonstrates that, when compared to the existing systems, our approach lowers the processing and communication overhead and provides the highest privacy preservation.