A Secure Softwarized Blockchain-based Federated Health Alliance for Next Generation IoT Networks

Abstract

Medical health care centres are seen as a potential paradigm for dealing with large amounts of data utilizing artificial intelligence (AI). The fast growth in the massive volume of data created by connected devices in the health paradigm brings up new opportunities for data sharing to improve the quality of service for developing applications. Traditional AI approaches frequently need centralized data gathering and model training within a single company, which is a significant flaw owing to the lack of privacy and security of raw data transfer through mobile networks. The leakage of sensitive data can result in substantial consequences for the providers, in addition to financial loss. For that, we first build a blockchain-enabled safe data sharing architecture for distributed multiple parties in this paper. Then, by adding suggested privacy-preserved federated learning, we turn the data-sharing problem into a machine-learning challenge. Recommendations are based on the user’s previous search values, and data privacy is protected by providing the data model rather than the raw figures. Finally, we include federated learning into the permissioned blockchain consensus process, allowing the consensus computing effort to be used for federated training. According to numerical findings generated from data sets, the suggested data sharing method delivers good accuracy, high efficiency, and increased security.