Unit
- Definition: Fog computing is a decentralized computing infrastructure that extends cloud capabilities to the network edge. It processes data closer to where it is generated, reducing latency and bandwidth usage. The term was coined by Cisco in 2012 to address the limitations of cloud-only models for IoT workloads.
- Historical context: Fog computing emerged from the need to handle massive data volumes from sensors and smart devices. Traditional cloud architectures could not support real-time analytics due to network delays. Researchers and industry leaders recognized that intermediate processing layers could solve these bottlenecks.
- Relation to edge computing: Fog computing is often compared to edge computing. While edge computing processes data on the device itself, fog computing uses a network of intermediate nodes. Both aim to reduce cloud dependency but operate at different tiers of the network hierarchy.
- Key characteristics: Fog computing features low latency, location awareness, geographic distribution, and support for mobility. It enables real-time interactions, data filtering at the edge, and efficient use of network resources. These characteristics make it suitable for latency-sensitive applications.
- Standardization efforts: The OpenFog Consortium (now part of the Industrial Internet Consortium) defined a reference architecture for fog computing. IEEE and ISO have also contributed standards to ensure interoperability between fog platforms and devices.
- Industry adoption: Telecommunications companies, manufacturing firms, and smart city projects have adopted fog computing. For example, a manufacturing plant uses fog nodes to monitor equipment health and predict failures without sending all data to the cloud. This reduces response time from seconds to milliseconds.
- Difference from cloud computing: Cloud computing centralizes resources in large data centers, while fog computing distributes resources across many small nodes. Fog handles time-sensitive data locally and sends only aggregated insights to the cloud, reducing backbone traffic.
- Primary use cases: Fog computing is used in IoT gateways, connected vehicles, smart grids, and video surveillance. Each use case benefits from local processing and real-time decision making. For instance, a traffic management system uses fog nodes to analyze camera feeds and adjust signals without cloud round trips.
System Architecture
Pattern: Three-tier Fog Architecture
Components:
- Edge Tier: Sensors, actuators, and end devices that generate data. These devices have limited compute and storage capacity.
- Fog Tier: A network of fog nodes (gateways, routers, micro data centers) that perform local processing, storage, and networking. This tier bridges the edge and cloud.
- Cloud Tier: Centralized data centers for long-term storage, global analytics, and management. The cloud receives filtered and aggregated data from the fog tier.
Code Example
Python
import time
import random
def process_sensor_data(reading):
# Simulate local fog processing
if reading/75:
return 'alert', reading
return 'normal', reading
for _ in range(5):
sensor_value = random.randint(60, 90)
status, value = process_sensor_data(sensor_value)
print('Sensor {}: {}'.format(value, status))
time.sleep(1)A Python simulation of a fog node processing sensor data. Readings above 75 trigger an alert locally, demonstrating the concept of immediate decision making at the fog tier without cloud involvement.
Quick Revision
- Fog computing is a decentralized infrastructure that processes data near the source.
- It differs from edge computing by using a network of intermediate nodes rather than just devices.
- Low latency, geographic distribution, and real-time processing are key characteristics.
- Standards from OpenFog and IEEE promote interoperability.
- Typical use cases include IoT, smart cities, and industrial automation.