Businesses of every size use the cloud for sharing resources and services, optimizing costs, and having the flexibility to scale computing power up or down quickly and easily.
At the same time, mobility has exploded, and in the US, the 3G and 4G networks aren’t that quick about transmitting information to and from the cloud.
Therefore, a simpler method of data storage and processing is beneficial for some local data rather than sending it all to the cloud for processing. Data might be processed directly on, say, a mobile device, or on hardware that resides between the mobile device and the internet. The place where that supplementary processing takes place has been dubbed “the fog.”
The Basics of Fog Computing
Fog computing, or “fogging,” is a computing model where data, processing, and apps are concentrated at the edge of the network rather than existing entirely (or almost entirely) in the cloud. The result is reduced service latency and a better end-user experience.
Fog devices, which may include mobile devices themselves, or routers, are geographically distributed across multiple platforms. The cloud may be “up there,” remote and abstract, but the fog is close to the “ground,” getting things done in physical proximity to devices. Rather than being powered by servers, the fog is powered by smaller, more dispersed computers, which in addition to mobile devices, may eventually be included in things like cars and appliances.
How Fog Computing Relates to the IoT
Cloud dependency in an era of “smart” devices can be far from ideal. With the emergence of the internet of things (IoT), jet engines, refrigerators, and smartwatches are being connected to wireless networks. All this data simply can’t be quickly transmitted to the cloud at the same rate it’s being generated, and as more objects become part of the IoT, there will be even more bottlenecking.
Fog computing basically says, let’s store and process at least some of this massive amount of data on devices themselves, or on devices that are between, say, the smartwatch and the cloud. The fog is not an alternative to the cloud, but a supplement to it. You can envision three layers of data crunching, with smart devices at “ground level,” fog computing hovering directly over the ground, and the cloud at the top.
Unlike the cloud, the fog is down here with us and our smart devices.
Applications Appropriate for Fog Computing
Fog computing addresses services and applications that require some processing, but that don’t necessarily need the cloud. Applications requiring very low latency (like processing jet engine data) would be appropriate for fog processing. Applications that are distributed geographically, like pipeline monitoring or environmental monitoring could also be appropriate for fog processing, as would large, distributed control systems like smart traffic lights.
Cisco, IBM, and Fog Computing
Cisco sells routers, and wants to turn them into small hubs for data gathering and for processing at least some of that data . Some data may be processed on the router itself, while other data may be uploaded to the cloud for more powerful processing. Routers, in other words, wouldn’t talk to the cloud unless it was necessary.
IBM’s goal, according to IBM executive Paul Brody, is to turn the traditional cloud-based internet inside out. In IBM’s research, the fog is made up of the devices that are already all around us, connected together. This would allow, say, smart devices to send software updates to each other rather than sending them through the cloud first.
Is Fog Computing Just a Bunch of Hype?
To some people, the fog is just another word for cloud plus IoT, and some say it’s an idea cooked up by Cisco to detract from that company’s lack of prominence in the cloud computing arena. In detractors’ view, the fog is a new name for an existing concept. Michael C. Daconta, VP of Advanced Technology at InCadence Strategic Solutions says that “Muddying the waters here helps no one except a company that is trying to insert itself into a crowded cloud conversation by trying to make its Internet routers into edge computing devices.”
Whether you see fog computing as a new paradigm or as made-up marketing hype, you’ll probably encounter the term over the next few years as the IoT gains traction. Fog computing takes some of the heavy lifting off regular cloud services by utilizing local resources for quicker and smoother processes, and whatever you want to call it, you can expect it to increase in importance as more objects become smart and connected.