【Author】
Alsamhi, Saeed Hamood; Almalki, Faris A.; Afghah, Fatemeh; Hawbani, Ammar; Shvetsov, Alexey, V; Lee, Brian; Song, Houbing
【Source】IEEE TRANSACTIONS ON GREEN COMMUNICATIONS AND NETWORKING
【Abstract】Edge Intelligence is an emerging technology which has attracted significant attention. It applies Artificial Intelligence (AI) closer to the network edge for supporting Beyond fifth Generation (B5G) needs. On the other hand, drones can be used as relay station (mobile drone edge intelligence) to gather data from smart environments. Federated Learning (FL) enables the drones to perform decentralized collaborative learning by developing local models, sharing the model parameters with neighbors and the centralized unit to improve global model accuracy in smart environments. However, drone edge intelligence faces challenges such as security and decentralization management, limiting its functions to support green smart environments. Blockchain is a promising technology that enables privacy-preserving data sharing in a distributed manner. There are several challenges that still need to be addressed in blockchain-based applications, such as scalability, energy efficiency, and transaction capacity. Motivated by the significance of FL and blockchain, this survey focuses on the synergy of FL and blockchain to enable drone edge intelligence for green sustainable environments. Moreover, we discuss the combination of FL and blockchain technological aspects, motivation, and framework for green smart environments. Finally, we discuss the challenges and opportunities, and future trends in this domain.
【Keywords】Drones; Blockchains; Green products; Security; Data models; Convergence; Biological system modeling; Smart environment; federated learning; blockchain; tethered drone; energy harvesting; sustainable; privacy; drone edge intelligence; green environment; energy efficiency; connectivity; QoS; B5G
【标题】无人机在 B5G 智能环境中的边缘智能：区块链和联邦学习协同作用
【摘要】边缘智能是一项备受关注的新兴技术。它将人工智能 (AI) 应用到更接近网络边缘的位置，以支持超越第五代 (B5G) 的需求。另一方面，无人机可以用作中继站（移动无人机边缘智能），从智能环境中收集数据。联邦学习 (FL) 使无人机能够通过开发本地模型、与邻居和集中单元共享模型参数来执行分散式协作学习，以提高智能环境中的全局模型准确性。然而，无人机边缘智能面临安全和分散管理等挑战，限制了其功能以支持绿色智能环境。区块链是一种很有前途的技术，可以以分布式方式实现隐私保护数据共享。在基于区块链的应用程序中仍有几个挑战需要解决，例如可扩展性、能源效率和交易容量。受 FL 和区块链重要性的启发，本次调查重点关注 FL 和区块链的协同作用，以使无人机边缘智能实现绿色可持续环境。此外，我们讨论了 FL 和区块链技术方面、动机和绿色智能环境框架的结合。最后，我们讨论了该领域的挑战和机遇，以及未来的趋势。
【关键词】无人机；区块链；绿色产品；安全;数据模型；收敛;生物系统建模；智能环境；联邦学习；区块链；系留无人机；能量收集;可持续的;隐私;无人机边缘智能；绿色环境；能源效率;连接性；服务质量； B5G
评论