While Fog computing may not entirely replace cloud, it does help it by reducing its data loads.

Due to the growth of IoT, the existing cloud network is unable to keep with the increasing data loads and processing demands, especially in real-time. The increase of consumer and enterprise devices connected to the IoT has put too much strain on cloud services from even the most cutting edge providers.  While it provides a centralized architecture, there is not enough bandwidth, and it costs too much money. Other problems faced in cloud services for IoT include high latency, vulnerability to cyber-attacks, lack of location awareness and downtime issues. To address these issues, companies are slowly moving to fog computing, which extends the cloud to be closer to the things that produce and act on IoT generated data. Here, the design uses local computing nodes (fog nodes) between the endpoints (e.g., sensors, cameras) and cloud data centers to gather, store and process data instead of using a remote cloud data center. It basically refers to a decentralized computing structure. Besides, its flexibility and ability to gather and process data from both the centralized cloud and the edge devices of a network make it one of the most resourceful and emerging technology to mitigate the information overload we face today.

Important Characteristics

Some of the main features of fog computing include low latency and location awareness, wide-spread geographical distribution, mobility and scalability to include many nodes. This is why fog computing systems are deployed very close to the end-users in a widely distributed manner. The fog computing nodes hosted possess sufficient computing power and storage capacity to handle the resource-intensive user requests. These fog nodes can also process tasks without third-party interference and collaboratively offer computational flexibility, better communication, and storage capacity in the IoT continuum. While enabling real-time interaction, fog based analytics can enhance conscious and responsive reproduced customer requirement relation. Since fog computing can run independently from the cloud as it has its servers, this ensures that the user gets continuous, uninterrupted services, even in the case of no network connectivity to the cloud. It also enhances the security of encrypted data as it stays closer to the end-user reducing exposure to hostile elements of the system it is deployed in. It also improves QoS and provides software and security updates to resource-constrained devices.

Fog vs. Edge

Sometimes, fog computing is frequently and often interchangeably used for the term edge computing. Though both provide the same functionalities in pushing both data and intelligence to analytic platforms near data sources, the key difference between edge computing and fog computing comes down to where the processing of that data takes place. In the case of edge computing, the processing of the data takes place close to where it is generated, like a programmable automation controller. On the contrary, in fog computing, data is processed within a fog node or IoT gateway, which is situated within the local area network (LAN) level of the network.

Common Applications

There is a diverse range of applications in fog computing. For instance, online streaming platforms can provide uninterrupted viewing services, owing to fog networking's capacity and elasticity for provisioning low latency, mobility, and location identifications, with real-time data analysis. Similarly, in the healthcare sector, we have a huge volume of patient data generated every day. Leveraging fog computing can reduce data transfers that take minutes and turn them into seconds. This is monumental in terms of patient care as it will increase the speed of service dramatically.

In the case of Automated Driving System (ADS), companies need to incorporate multiple advanced technologies like multi-modal sensors, computer vision, artificial intelligence and machine learning, etc. These help the system perform data fusion, image analysis, mapping, and predictions to determine the best action and controls for the drive-train. So, having a fog computing environment will enable communications for all of these data functions both at the edge (in the car) and to its endpoint (the manufacturer).