What is Fog Computing?

Fog computing is an infrastructure for computing that is situated between the edge and the centralized cloud. It is similar to edge computing, but fog computing offers more computing power than edge and does so faster than cloud as it is nearer to where the data originates and is acted upon.  

The terms edge and fog are often used interchangeably as they employ similar attributes of being closer to the source of data creation. Fog computing, or fogging, offers benefits such as improved efficiency through reduced latency, the time taken to send, analyze and receive data, and added security. 

Fog is useful where edge computing, which operates usually within a device itself, is not powerful enough to do advanced analysis and/or machine learning tasks with data collected.

Fog Computing v Cloud Computing

While cloud has the resources to do extremely complex computational analysis, it is often far away from the data source meaning such processing of data does not take place in a timely manner – i.e. it takes too long. Fog computing enables data analysis and response in real time, much faster than cloud. 

Another potential concern with cloud is that connecting to and transmitting data across the internet can throw up privacy, security, and legal concerns particularly regarding critical, sensitive data that may run into regulatory problems in different countries, for example. As fog computing occurs near the source of data, those concerns are mitigated.

Fog computing v Edge Computing

In essence, the main difference between fog and edge computing is where the data and computation power are located. Purely fog computing takes place at the local area network (LAN) site - the data is sent from endpoints to a fog gateway, which is a network node that connects different networks with different transmission protocols, and then sent to sources for processing and return transmission.

Examples of Fog Computing 

Fogging applications today are often employed with systems such as smart grids, smart buildings and smart cities, vehicle networks, and software-defined networks. 

One example of fogging use is traffic control where real-time analysis of traffic data is collated and processed by smart city systems, allowing traffic signals to respond appropriately and instantly to variable conditions. Such systems are subject to far less latency in fog computing, simply because they are closer to the source of data.  

Increasingly, the development of autonomous vehicles is seen to utilize fog computing as the amount of data generated by an autonomous car is vast but needs to be acted upon in a real-time manner. The amount of data is too much to be processed onboard by the vehicle itself but the cloud is too far away to allow instant decision-making and in addition requires continual connection to the internet. By using fog computing, the volume of data can be analyzed and acted upon virtually instantly without the time lag of cloud but with more power than edge. 

Fog Computing in Conclusion

Effectively, fog computing is where a number of nodes receive information from devices, be they Internet of Things (IoT) devices or, for instance, factory production line sensors, and act upon it in real time. It is more powerful than edge computing alone and provides mission-critical analysis faster than cloud. The ability to conduct data analysis in real-time means faster alerts to potential failures and less likelihood of time lost in production process breakdowns, for example. Sometimes these fog nodes send summaries of data analysis to the cloud where it can be further analyzed to enable predictive decision making regarding various aspects of the device or systems, such as functionality and system health. 

Fogging needs high-speed connectivity between devices and the nodes collecting data to enable the practically instant processing such devices and systems require. These can be wired, in the case of an IoT sensor on a factory production line for instance, or they can be digitally linked, perhaps via 5G, such as autonomous vehicles or wind turbines in fairly remote locations. 

Fog computing also uses less network bandwidth. It is a fact that many analysis tasks, even critical ones, do not require the scale that cloud-based storage and processing offers. In addition, vast amounts of data are produced by connected devices. Fog computing cuts out the need to move the majority of this huge amount of data to the cloud, thereby saving bandwidth for other more mission critical tasks. 

Fog computing also cuts operating costs because the majority of data is processed locally. Alongside that, as IoT devices are regularly used in difficult environmental conditions and/or during emergencies, fog computing boosts reliability under such conditions, reducing the data transmission burden. 

While fog computing and related applications are complex issues, here at Techbuyer we have the expertise needed to help you navigate those issues to best optimise your organization’s compute requirements. To find out more drop us a line or give us a call.