If there was an official model for the internet, it would be this: "run before you can walk." Cloud computing is barely in its adolescence—but forward-thinkers are already talking about its descendent. Fog computing may be a necessary reaction to the proliferation of devices at the edge of networks, but the philosophy makes a lot of assumptions about how networks will work in the future. If networks begin lurching towards the fog before we've even fully sorted out the cloud, will you be prepared to support the move?
The Cloud Isn’t Broken, So Why Fix It?
Everyone knows how the cloud computing model works: Edge devices such as laptops, desktops, IP phones, and other terminals are relatively constrained by limitations on storage and processing power. Cloud computing removes this constraint—you put all the processing power and storage into a data center somewhere else, and then pipe that over to a given application when needed. All of a sudden, your only limits are the size of your bandwidth and the size of your data center.
This model works well, and enables hugely successful business models like SaaS and Big Data—but then there's the Internet of Things (IoT). All of a sudden, you have devices connected to the cloud with the minimum amount of processing power available. This inevitably leads to problems as these devices aren't even smart enough to register when they're being taken over by unauthorized third parties. By the time the controller in the cloud realizes what's going on, you've got a botnet made out of WiFi teakettles.
Fog Networking: What Is This?
IoT represents an expanding network of relatively dumb devices, and fog networking proposes to fix the problems that this presents. Security concerns aside, the crowd of devices on the edge are also eating up the last big constraint on cloud performance: bandwidth. Too many devices are delivering too-large amounts of data to processing centers that may not even be on the same continent. Left unchecked, this swarm of input will eventually slow the cloud to a crawl.
Fog computing aims to nip this problem in the bud by moving computing resources closer to IoT sensors. This approach could take a variety of forms. For example, one could:
- Use the flexibility of SD-WAN to route traffic from IoT devices to whatever data center happens to be closest.
- Increase the processing power of IoT devices by a small amount, and create an intelligent mesh network that processes data by minimizing connections to the cloud.
- Anchor IoT devices with an appliance that performs on-board data-processing and analysis.
Service Providers: How Will You See Through the Fog?
It’s only a matter of time before someone will come up with a fog networking application that will apply to VoIP. Maybe it's an IP phone with enough embedded processing power to make call routing decisions without first connecting to a PBX in the cloud. Assuming your business model survives this transition, the problem becomes this: are you prepared to troubleshoot this new networking model once it inevitably breaks?
Edgewater Networks offers a platform-agnostic solution to problems such as the emergence of new service delivery models for platforms such as UC and VoIP. Our Edgeview Service Control Center already allows service providers to diagnose errant connections without resorting to a truck roll—by the time the network becomes drenched in fog, we'll be ready to see through it. For more information on how Edgewater Networks can help service providers adapt to a changing landscape, check out our white paper, "Digital Transformation at the Edge—How Service Providers Can Stay Ahead of Disruption."