1) Description about Fog computing
Fog computing operates at ends of the network, not like cloud computing which operates from a central location. It places some resources and processes on edges of the cloud. It is also known as edge computing. Due to some limitations related to technical and infrastructure aspects, there are some services and applications which could not fit in a paradigm of cloud, to address them, a paradigm of cloud computing to an edge of the network is extended by fog computing.
2) Difference between Fog computing and cloud computing
In Cloud Computing, there are a different set of machines distributed and running at various locations while connected to a single hub service or a network.
In Fog computing, one or more than one collaborative multitude of customers carried storage, configuration, measurement, communication, control, and management in a substantial amount.
3) Applications of fog computing
There are some tech giants such as IBM which is driving force behind fog computing. These days in offices, the data center is a common thing or you can see hundreds of devices are connected and communicated to each other and it is expecting that in next few years this number could rise to thousands or lakhs. There are possibilities that direct user-end computing and communication will soon become more relevant.
Here are some practical examples for you to understand where fog computing can be applied:
- It permits prompt interaction between machine and humans, between machine and machine with the cooperation of cloud.
- The activities which are going in the cities, on their levels, Sensor data can be obtained by fog computing and then it integrates all network entities which are working independently within those cities.
- As per reports of Markets, it is predicted that by 2017, cloud computing market in the field of healthcare will reach $5.4 billion and the same thing on a local level will be permitted by fog computing.
Fog computing laid focus on a big scale on processes which brings us closer to the location of data as much as possible. Invaluable data must be stopped from reaching to a cloud and only data which has some worth or has some value should reach to networks of cloud computing.
Models which helps us to learn machines, through them fog computing is done in the best manner which gets training on a small part of data in the cloud. If this model finds suitable, then it is sent nearer to the devices. On these devices, there are some algorithms such as decision trees and fuzzy logic can be used on a local level in order to take a decision which is less expensive as compared setting up of an infrastructure in a cloud which is required to deal with data acquired from tens of thousands of devices.
4) What is next for fog computing?
If you want services on an immediate basis, fog computing is a good way. It can also be used in place of the internet as a speed of net is dependent on carriers.
For future, large companies such as Facebook, Google are looking for alternative methods such as drones and balloons to access the web in case any network issue occurs. And small-scale organizations are planning to develop a fog of devices which are currently present around them so as to establish close and prompt connections by which resources can be computed.
Aggregated and centralized cloud computing will certainly have a place but as sensors moved to things and data also grows enormously, thus in order to host applications, a new and a fresh approach is required. The right approach for hosting fresh and new applications which could make use of existing devices is Fog computing.
Although the relevance of center is not reduced because of the movement to edge. And on other hands, it reflects that to expand computer architecture, datacentre must be a great nucleus.