Publication

Abstract : Modern Internet of Things (IoT) applications generate enormous amounts of data. To prototype the incoming data from such applications, data-driven machine learning has emerged as a viable method that can help develop precise and reliable statistical models. Existing data is not effectively utilized by machine learning as it is stored in data silos, and technology businesses are now required to treat user data carefully in accordance with user-privacy legislation in many regions of the world. It is self-evident that the samples in the typical machine learning centralized server paradigm have different probability distributions of data supplied. As a result, the common model fails to personalize. Without sufficient data, machine learning is unlikely to reach its full potential. Hence, the new distributed paradigm, federated learning, has gained popularity that supports collaborative learning while preserving privacy. IoT applications with cryptography characteristics will be capable of storing and transmitting data securely over networks by maintaining data consistency. Therefore, federated learning and blockchain integration are 238particularly beneficial for IoT applications that manage sensitive data such as healthcare. Although several studies have focused on various applications of blockchain technology and federated learning, a thorough examination of these technologies in edge-fog-cloud-based IoT computing systems and healthcare applications has not yet been performed. The basic architecture, structure, types, functions, and characteristics of federated learning and blockchain are addressed at the outset of this survey article. This article also considers the wide range of applications to which federated learning and blockchain have been implemented, including edge, fog, cloud, and IoT computing paradigms. Lastly, it evaluates the different implementations of federated learning in healthcare applications.