Data Analytics for Fog Computing by Distributed Online Learning with Asynchronous Update

Abstract

Fog computing extends the cloud computing paradigm by allocating substantial portions of computations and services towards the edge of a network, and is, therefore, particularly suitable for large-scale, geo-distributed, and data-intensive applications. As the popularity of fog applications increases, there is a demand for the development of smart data analytic tools, which can process massive data streams in an efficient manner. To satisfy such requirements, we propose a system in which data streams generated from distributed sources are digested almost locally, whereas a relatively small amount of distilled information is converged to a center. The center extracts knowledge from the collected information, and shares it across all subordinates to boost their performances. Upon the proposed system, we devise a distributed machine learning algorithm using the online learning approach, which is well known for its high efficiency and innate ability to cope with streaming data. An asynchronous update strategy with rigorous theoretical support is applied to enhance the system robustness. Experimental results demonstrate that the proposed method is comparable with a model trained over a centralized platform in terms of the classification accuracy, whereas the efficiency and scalability of the overall system are improved.

Ambient Backscatter-Assisted Wireless-Powered Relaying

“…Proof. Recall that, from the proof of Theorem 2, we have obtained the probability of R selecting the ABR mode and WPR mode under ETCP at the steady states in (42) and (43), respectively. According to the mode selection criteria of ETCP described in Section II, the ergodic capacity of the hybrid relaying with ETCP at the steady states can be expressed as…”

“…, C ABR , and C WPR obtained in (42), (43), (25), and (28), respectively, into (32), we have the expression of C ETCP HR in (31).…”

“…The advantage of the hybrid relay lies in the fact that it can select the more suitable operational mode to achieve better performance under different network conditions. However, as the hybrid relay is a wireless-powered device, any mode selection protocols that have heavy computational complexity, e.g., based on online optimizations [43], [44], are not applicable. For implementation practicality, low complexity mode selection based on obtainable information and limited communication overhead needs to be devised.…”

A Bandit Approach for Mode Selection in Ambient Backscatter-Assisted Wireless-Powered Relaying

“…A promising future direction is to design bandit mode selection approaches for intelligent reconfigurable surface [17], [18] to assist relaying. Our system model can also be extended to case with multiple hybrid relays which can integrate distributed online learning [19], [20] with bandit for mode selection.…”

Improving IoT Analytics through Selective Edge Execution

“…For instance, [21], [22] tailor deep neural networks for execution in mobile devices, while [23] and [24] minimize the execution time for known system parameters and task loads. Finally, [25]- [27] leverage the edge architecture to effectively execute analytics for IoT devices. The plethora of such system proposals, underlines the necessity for our online decision framework that provides optimal execution of analytics.…”