Access and Feeds

Fog and Edge Computing: Network Architectures that Decentralize IoT Data Collection

By Dick Weisinger

Many applications have been built with a client-server model that centralizes most of the processing and data storage on a central server.  But if huge numbers of clients are added to the network, the model breaks down when the server becomes overloaded and overtasked with interactions from the many clients.  Communications back and forth of large amounts of data from client devices can also put a strain on available network bandwidth.  It’s possible to bulk up server capabilities by scaling the server up or out, but that often comes at considerable cost and complexity.

An Internet of Things (IoT) architecture typically involves very many devices that capturing data.  To minimize huge bandwidths of raw data streaming from devices back to a central server, there are two designs to minimize loads put on a central server: fog computing and edge computing.  Both of these approaches decentralize computing.

In an edge computing model, the devices themselves that collect data are powerful enough to be able to process the data.  Often these devices will be built with programmable automation controllers (PACs).

Fog computing puts processing nodes out in the field close to the devices.  These fog nodes serve as a gateway servers to a group of IoT devices.

Brent Hodges, head of IoT planning and product strategy at Dell, said that “we work to make sure fog computing isn’t just about the edge. It’s about the ‘and’ — taking elements from the cloud and getting closer to where computing is happening.”

Mohannad Abuissa, head of sales engineering at Cisco Middle East, said that “analyzing IoT data close to where it is collected minimizes latency. Along with the cloud, fog computing is assisting in accelerating the adoption of the IoT in the enterprise. It offloads gigabytes of network traffic from the core network, and keeps sensitive data inside the network.”

The global fog computing market is now only about a $10 million market, but it is expected to grow by about 55 percent annually through 2022.

 

 

 

 

Digg This
Reddit This
Stumble Now!
Buzz This
Vote on DZone
Share on Facebook
Bookmark this on Delicious
Kick It on DotNetKicks.com
Shout it
Share on LinkedIn
Bookmark this on Technorati
Post on Twitter
Google Buzz (aka. Google Reader)

Leave a Reply

Your email address will not be published. Required fields are marked *

*