Edge computing brings computing and storage resources closer to (mobile) end users and data sources, thus bypassing expensive and slow links to distant cloud computing infrastructures. Often leveraged opportunistically, these heterogeneous resources can be used to offload data and computations, enabling upcoming demanding applications such as augmented reality and autonomous driving. Research in this direction has addressed various challenges, from architectural concerns to runtime optimizations. As of today, however, we lack a widespread availability of edge computing-partly because it remains unclear which of the promised benefits of edge computing are relevant for what types of applications. This article provides a comprehensive snapshot of the current edge computing landscape, with a focus on the application perspective. We outline the characteristics of edge computing and its postulated benefits and drawbacks. To understand the functional composition of applications, we first define common application components that are relevant w.r.t. edge computing. We then present a classification of proposed use cases and analyze them according to their expected benefits from edge computing and which components they use. Furthermore, we illustrate existing products and industry solutions that have recently surfaced and outline future research challenges. INDEX TERMS Edge computing, heterogeneous networks, next generation networking mobile applications, Internet of Things, ubiquitous computing.

No abstract

No abstract

Federated learning enables multiple users to build a joint model by sharing their model updates (gradients), while their raw data remains local on their devices. In contrast to the common belief that this provides privacy benefits, we here add to the very recent results on privacy risks when sharing gradients. Specifically, we investigate Label Leakage from Gradients (LLG), a novel attack to extract the labels of the users’ training data from their shared gradients. The attack exploits the direction and magnitude of gradients to determine the presence or absence of any label. LLG is simple yet effective, capable of leaking potential sensitive information represented by labels, and scales well to arbitrary batch sizes and multiple classes. We mathematically and empirically demonstrate the validity of the attack under different settings. Moreover, empirical results show that LLG successfully extracts labels with high accuracy at the early stages of model training. We also discuss different defense mechanisms against such leakage. Our findings suggest that gradient compression is a practical technique to mitigate the attack.

Federated learning enables multiple users to build a joint model by sharing their model updates (gradients), while their raw data remains local on their devices. In contrast to the common belief that this provides privacy benefits, we here add to the very recent results on privacy risks when sharing gradients. Specifically, we propose Label Leakage from Gradients (LLG), a novel attack to extract the labels of the users' training data from their shared gradients. The attack exploits the direction and magnitude of gradients to determine the presence or absence of any label. LLG is simple yet effective, capable of leaking potential sensitive information represented by labels, and scales well to arbitrary batch sizes and multiple classes. We empirically and mathematically demonstrate the validity of our attack under different settings. Moreover, empirical results show that LLG successfully extracts labels with high accuracy at the early stages of model training. We also discuss different defense mechanisms against such leakage. Our findings suggest that gradient compression is a practical technique to prevent our attack.