Fog computing has emerged as an essential alternative to the cloud. Fog computing is the future as it is nearer to the edge where actually the IOT devices and sensors are located. A Fog Server or Fog Node is located near to the IOT devices, connecting directly (wired or wireless) to them. The Fog Server has a functionality of fast accessibility to the data arising out of IOT devices or sensors, as against cloud server which may be located in data centers (near core Network Centers) located far away from the edge resulting in extreme delays in network transmission and latency, especially when the data is large volume as stream (or ‘Big Data’) arising out of IOT devices or sensors including cameras, etc. Real time response after completing the necessary Analytics on the data generated by IOT devices and sensors becomes critically essential for meeting the real time response requirements of critical applications such as in health care and transportation. What are the relevant techniques for Fog Analytics? In this paper we provide a brief survey of Fog Analytics techniques in stream data analytics, machine learning, deep learning techniques and also game theoretical adversarial learning.
- Fog computing