Logo
ScienceStack
Table of Contents
Fetching ...
Sign In

Distributed Edge Analytics in Edge-Fog-Cloud Continuum

Satish Narayana Srirama

TL;DR

The paper addresses the latency, bandwidth, and privacy challenges of Cloud-centric IoT by advocating distributed edge analytics across the edge-fog-cloud continuum. It presents three complementary viewpoints—serverless data pipelines (SDP) implemented via MQTT, distributed computing with the CANTO framework, and federated learning with the FIDEL platform—each demonstrated through real-world case studies. The SDP case shows event-driven processing along the continuum, the CANTO-based distributed learning on fog clusters yields competitive accuracy for forest-fire prediction, and the FL setup demonstrates privacy-preserving model training across edge devices with improving accuracy over rounds. Together, these results highlight practical, scalable approaches for enabling edge analytics with reduced latency and enhanced privacy in IoT deployments.

Abstract

To address the increased latency, network load and compromised privacy issues associated with the Cloud-centric IoT applications, fog computing has emerged. Fog computing utilizes the proximal computational and storage devices, for sensor data analytics. The edge-fog-cloud continuum thus provides significant edge analytics capabilities for realizing interesting IoT applications. While edge analytics tasks are usually performed on a single node, distributed edge analytics proposes utilizing multiple nodes from the continuum, concurrently. This paper discusses and demonstrates distributed edge analytics from three different perspectives; serverless data pipelines (SDP), distributed computing and edge analytics, and federated learning, with our frameworks, MQTT based SDP, CANTO and FIDEL, respectively. The results produced in the paper, through different case studies, show the feasibility of performing distributed edge analytics following the three approaches, across the continuum.

Distributed Edge Analytics in Edge-Fog-Cloud Continuum

TL;DR

The paper addresses the latency, bandwidth, and privacy challenges of Cloud-centric IoT by advocating distributed edge analytics across the edge-fog-cloud continuum. It presents three complementary viewpoints—serverless data pipelines (SDP) implemented via MQTT, distributed computing with the CANTO framework, and federated learning with the FIDEL platform—each demonstrated through real-world case studies. The SDP case shows event-driven processing along the continuum, the CANTO-based distributed learning on fog clusters yields competitive accuracy for forest-fire prediction, and the FL setup demonstrates privacy-preserving model training across edge devices with improving accuracy over rounds. Together, these results highlight practical, scalable approaches for enabling edge analytics with reduced latency and enhanced privacy in IoT deployments.

Abstract

To address the increased latency, network load and compromised privacy issues associated with the Cloud-centric IoT applications, fog computing has emerged. Fog computing utilizes the proximal computational and storage devices, for sensor data analytics. The edge-fog-cloud continuum thus provides significant edge analytics capabilities for realizing interesting IoT applications. While edge analytics tasks are usually performed on a single node, distributed edge analytics proposes utilizing multiple nodes from the continuum, concurrently. This paper discusses and demonstrates distributed edge analytics from three different perspectives; serverless data pipelines (SDP), distributed computing and edge analytics, and federated learning, with our frameworks, MQTT based SDP, CANTO and FIDEL, respectively. The results produced in the paper, through different case studies, show the feasibility of performing distributed edge analytics following the three approaches, across the continuum.
Paper Structure (6 sections, 4 figures)

This paper contains 6 sections, 4 figures.

Table of Contents

  1. Introduction
  2. State-of-the-art in data management in the edge-fog-cloud continuum and edge analytics
  3. Serverless data pipelines in edge-fog-cloud continuum
  4. Distributed computing and data analytics on the fog nodes
  5. Federated learning
  6. CONCLUSION

Figures (4)

  • Figure 1: Serverless data pipeline for IIoT surveillance
  • Figure 2: (a) SDP completion time for each image (b) SDP execution time per image
  • Figure 3: Federated learning on the edge-fog-cloud continuum
  • Figure 4: FL accuracy over multiple rounds (a) Across all devices (b) On the test data with aggregated model on server