Federated Learning For IoT Devices in Smart Cities: A Particle Swarm Optimation-Based Approach

Authors:
Publisher website
Google Scholar
Abstract
Smart cities witnessed significant advancements in integrating Internet of Things (IoT) wheareas the IoT-generated big data in smart cities raised concerns about privacy and security. This research paper addresses privacy and security concerns in utilizing IoT-generated big data in smart cities. It proposes a novel approach that combines federated learning (FL) with the particle swarm optimization (PSO) algorithm to enhance prediction accuracy while ensuring data privacy. The FL approach distributes the learning process across IoT devices preserving data privacy and leveraging collective intelligence. Extensive evaluation using privacy and security metrics tailored for smart cities which demonstrates the effectiveness of the proposed approach. It achieves high classification accuracy, maintains data privacy and mitigates privacy and security risks. This research contributes to knowledge on privacy and security in big data analysis by providing insights for smart cities and IoT-generated data. The findings have important implications for responsible development and use of IoT-generated data in smart cities.

This publication has 8 references indexed in Scilit: