Fog Computing

The primary aim is to solve the problems faced by cloud computing during IoT data processing. it is an intermediate layer between cloud and devices. 
Fog computing or fogging is coined by CISCO. The idea of fog computing is to extend the cloud nearer to the IoT devices.

Why Fog Computing
The ability of the current cloud model is insufficient to handle the requirements of IoT.
Issues resolved by fog are:
 a) Volume
 b) Latency
 c) Bandwidth

The architecture of Fog:
Cloud services are extended to IoT devices through fog
Fog is a layer between cloud and IoT devices
many fog nodes can be present
Sensor data are processed in the fog before it is sent to the cloud
Reduces latency, save bandwidth and save the storage of the cloud

Working of Fog
 Three types of data
 1: Very time‐sensitive data
      Data which should be analyzed within fraction of a second
      Analyze at the nearest node itself
     Sends the decision or action to the devices
    Sends and stores the summary to cloud for future analysis

2: Less time‐sensitive data
    Data which can be analyzed after seconds or minutes
   Are sent to the aggregate node for analysis
  After analysis, the aggregate node send the decision or action to   the device through the nearest node
 The aggregate node sends the summary to cloud for storage and
 future analysis.

3: Data which are not time‐sensitive
     Data which can be wait for hours, days, weeks
    Sent to cloud for storage and future analysis.
    Those summaries from fog nodes can be considered as less time
     sensitive data.

Fog nodes works according to the type of data they receive. An IoT application should be installed to each fog nodes