Federated Learning Meets Blockchain in Edge Computing: Opportunities and\n Challenges

Mobile edge computing (MEC) has been envisioned as a promising paradigm to\nhandle the massive volume of data generated from ubiquitous mobile devices for\nenabling intelligent services with the help of artificial intelligence (AI).\nTraditionally, AI techniques often require centralized data collection and\ntraining in a single entity, e.g., an MEC server, which is now becoming a weak\npoint due to data privacy concerns and high data communication overheads. In\nthis context, federated learning (FL) has been proposed to provide\ncollaborative data training solutions, by coordinating multiple mobile devices\nto train a shared AI model without exposing their data, which enjoys\nconsiderable privacy enhancement. To improve the security and scalability of FL\nimplementation, blockchain as a ledger technology is attractive for realizing\ndecentralized FL training without the need for any central server.\nParticularly, the integration of FL and blockchain leads to a new paradigm,\ncalled FLchain, which potentially transforms intelligent MEC networks into\ndecentralized, secure, and privacy-enhancing systems. This article presents an\noverview of the fundamental concepts and explores the opportunities of FLchain\nin MEC networks. We identify several main topics in FLchain design, including\ncommunication cost, resource allocation, incentive mechanism, security and\nprivacy protection. The key solutions for FLchain design are provided, and the\nlessons learned as well as the outlooks are also discussed. Then, we\ninvestigate the applications of FLchain in popular MEC domains, such as edge\ndata sharing, edge content caching and edge crowdsensing. Finally, important\nresearch challenges and future directions are also highlighted.\n