Combining Federated Learning and Edge Computing toward Ubiquitous Intelligence: Challenges, Recent Advances, and Future Directions

Abstract

<p>Full leverage of the huge volume of data generated on a large number of user devices for providing intelligent services in the 6G network calls for Ubiquitous Intelligence (UI). A key to developing UI lies in the involvement of the large number of network devices, which contribute their data to collaborative Machine Learning (ML) and provide their computational resources to support the learning process. Federated Learning (FL) is a new ML method that enables data owners to collaborate in model training without exposing private data, which allows user devices to contribute their data to developing UI. Edge computing deploys cloud-like capabilities at the network edge, which enables network devices to offer their computational resources for supporting FL. Therefore, a combination of FL and edge computing may greatly facilitate the development of ubiquitous intelligence. </p> <p>In this article, we present a comprehensive survey of the recent developments in technologies for combining FL and edge computing with a holistic vision across the fields of FL and edge computing. We conduct our survey from both the perspective of an FL framework deployed in an edge computing environment (<em>FL in Edge</em>) and the perspective of an edge computing system providing a platform for supporting FL (<em>Edge for FL</em>). From the <em>FL in Edge</em> perspective, we first identify the main challenges to FL in edge computing and then survey the representative technical strategies for addressing the challenges. From the <em>Edge for FL</em> perspective, we first analyze the key requirements for edge computing to support FL and then review the recent advances in edge computing technologies that may be exploited to meet the requirements. Then we discuss open problems and identify some possible directions for future research on combining FL and edge computing, with the hope of arousing the research community’s interest in this emerging and exciting interdisciplinary field. </p>

Energy Conservation in Passive Optical Networks: A Tutorial and Survey

“…Later, based on APON, ITU-T introduced BPON with higher transmission capacity than APON (BPON provides 1244. 16 Mbps and 622 Mbps data rate for downlink and uplink communication, respectively [83]).…”

“…Furthermore, an important challenge is how to ensure uninterrupted power supply to ONUs to facilitate flawless c omputing o peration. A nother i ssue needs to be tackled in built-in PON edge computing is security and privacy protection (this has been an important research challenge in the domain of edge computing [16]).…”

“…PON has also been widely adopted not only to connect the residential and business users but also to serve wireless access segments, such as fourth-generation (4G) and fifth-generation (5G), as a backhaul [14], [15]. The upcoming sixthgeneration (6G), which envisions to offer high data and ultralow latency mobile Internet and ubiquitous intelligence [16], is expected to use PON as an optical transport solution as a backhaul [17]. Despite the fact that PON is regarded as an energy-efficient access network technology, PON would be one of the major contributors for energy consumption in the access network segment due to its high penetration rate [18].…”

Layer-Wise Personalized Federated Learning for Mobile Traffic Prediction

“…Furthermore, due to the advancements in edge computing technology to support high QoS requirements of specific use cases in mobile networks, powerful data centers are deployed across multiple regions. This distributed paradigm has paved the way for FL [8], where data can be trained locally in each region, and effectively learned models can be shared. Given these multifaceted advantages, FL is expected to play a pivotal role in next-generation communications [9].…”