Towards Intelligent and Secure Healthcare: A Survey on the Role of Edge Computing, Federated Learning and Homomorphic Encryption

This research explores solely the integration of Homomorphic encryption with Federated Learning and Edge Computing within the healthcare domain. Edge Computing focuses on processing the data at the edge of the network, rather than sending the data to the centralized cloud to process the data and send the processed data back to the device, this approach will thereby offer reduced latency and enhanced real-time data handling, which is crucial for most healthcare devices which are IoT based devices. Federated Learning works based on Federation where data is shared between two parties. Here we focus on Machine Learning in a distributed approach, where the models can be trained across multiple decentralized devices while maintaining data privacy. One may argue that security is the major flaw in this approach, and this is where we can use an encryption technique called homomorphic encryption. It allows computation to be performed on the encrypted data itself, so there is no need for data decryption throughout the data processing lifecycle so the security and reliability on the central servers and the threat of data breaches are reduced. The synergy between these technologies is highlighted in various applications, including remote health monitoring, smart hospitals, and secure data management. Choosing the right framework is key to the implementation of federated learning in the healthcare domain. The research provides a detailed comparison of the most popular frameworks in terms of data privacy, scalability, performance, and computational complexity.