In the traditional cloud architecture, data needs to be uploaded to the cloud for processing, bringing delays in transmission and response. Edge network emerges as the times require. Data processing on the edge nodes can reduce the delay of data transmission and improve the response speed. In recent years, the need for artificial intelligence of edge network has been proposed. However, the data of a single, individual edge node is limited and does not satisfy the conditions of machine learning. Therefore, performing edge network machine learning under the premise of data confidentiality became a research hotspot. This paper proposes a Privacy-Preserving Asynchronous Federated Learning Mechanism for Edge Network Computing (PAFLM), which can allow multiple edge nodes to achieve more efficient federated learning without sharing their private data. Compared with the traditional distributed learning, the proposed method compresses the communications between nodes and parameter server during the training process without affecting the accuracy. Moreover, it allows the node to join or quit in any process of learning, which can be suitable to the scene with highly mobile edge devices.INDEX TERMS Federated learning, edge computing, privacy preservation, asynchronous distributed network, gradient compression.

No abstract

Understanding mobile app usage has become instrumental to service providers to optimize their online services. Meanwhile, there is a growing privacy concern that users' app usage may uniquely reveal who they are. In this paper, we seek to understand how likely a user can be uniquely re-identified in the crowd by the apps she uses. We systematically quantify the uniqueness of app usage via large-scale empirical measurements. By collaborating with a major cellular network provider, we obtained a city-scale anonymized dataset on mobile app traffic (1.37 million users, 2000 apps, 9.4 billion network connection records). Through extensive analysis, we show that the set of apps that a user has installed is already highly unique. For users with more than 10 apps, 88% of them can be uniquely re-identified by 4 random apps. The uniqueness level is even higher if we consider when and where the apps are used. We also observe that user attributes (e.g., gender, social activity, and mobility patterns) all have an impact on the uniqueness of app usage. Our work takes the first step towards understanding the unique app usage patterns for a large user population, paving the way for further research to develop privacy-protection techniques and building personalized online services.

The advent of publishing anonymized call detail records opens the door for temporal and spatial human dynamics studies. Such studies, besides being useful for creating universal models for mobility patterns, could be also used for creating new socio-economic proxy indicators that will not rely only on the local or state institutions. In this paper, from the frequency of calls at different times of the day, in different small regional units (sub-prefectures) in Côte d'Ivoire, we infer users' home and work sub-prefectures. This division of users enables us to analyze different mobility and calling patterns for the different regions. We then compare how those patterns correlate to the data from other sources, such as: news for particular events in the given period, census data, economic activity, poverty index, power plants and energy grid data. Our results show high correlation in many of the cases revealing the diversity of socio-economic insights that can be inferred using only mobile phone call data. The methods and the results may be particularly relevant to policy-makers engaged in poverty reduction initiatives as they can provide an affordable tool in the context of resource-constrained developing economies, such as Côte d'Ivoire's.

We analyze a class of distributed quantized consensus algorithms for arbitrary networks. In the initial setting, each node in the network has an integer value. Nodes exchange their current estimate of the mean value in the network, and then update their estimation by communicating with their neighbors in a limited capacity channel in an asynchronous clock setting. Eventually, all nodes reach consensus with quantized precision. We start the analysis with a special case of a distributed binary voting algorithm, then proceed to the expected convergence time for the general quantized consensus algorithm proposed by Kashyap et al. We use the theory of electric networks, random walks, and couplings of Markov chains to derive an O(N 3 log N ) upper bound for the expected convergence time on an arbitrary graph of size N , improving on the state of art bound of O(N 4 log N ) for binary consensus and O(N 5 ) for quantized consensus algorithms. Our result is not dependent on graph topology. Simulations on special graphs such as star networks, line graphs, lollipop graphs, and Erdös-Rényi random graphs are performed to validate the analysis.This work has applications to load balancing, coordination of autonomous agents, estimation and detection, decision-making networks, peer-to-peer systems, etc.

No abstract

No abstract

No abstract