• 07/09/2014
    8:00 AM
  • Rating: 
    0 votes
    Vote up!
    Vote down!

IoT: The Network Impact

The Internet of Things will require a network that can handle increased demand for data analytics, agility, and security.

There can be no Internet of Things (IoT) without the network to support it. Sensors and gadgets will gather increasingly vast amounts of data. But the Internet of Things is about more than just gadgets and displays; the amount of data gathered will seriously impact the network, and the networking industry needs to evaluate possible implications.

Three areas of the IoT that will impact the network are data analytics, the need for network agility, and security. Let's take a closer look at these three areas.

Data analytics
Data without analytics is relatively useless. The influx of sensors will create vast amounts of data that will need to be processed. For manufacturers, post-sale service or warranties can be continuously tracked in real-time using machine-to-machine sensors to identify malfunctions or warranty issues. Real-time promotions can be sent by analyzing sensor data and customers' buying preferences.

These newfound capabilities made available across industries will not only increase the amount of data, but also significantly increase the demand for business intelligence. It's important to think about the backend implications this will present to the network.

The sheer volume of data will increase the drive toward a cloud-based data center as moving data efficiently to the cloud and extracting intelligence become very critical tasks.

Agile networking
The IoT will impact everyone at the professional and individual levels; whole industries will leverage the availability of sensors and machine-to-machine communication. In agriculture, irrigation systems will function based on a multitude of inputs, including weather forecasts and data from moisture sensors. Manufacturing plants will be totally wired on sensor networks, as will oil drilling equipment. Drilling times will be reduced by the use of advanced analytics that can predict conditions and improve operation based on previous events.

The sheer volume of data created by the IoT will have unfathomable impact on the networking systems used today. Deep analytics will require distributed datacenters and real-time response to events. Fast, agile networks are crucial to enable the real-time analysis of sensor data. Given these requirements, it is very unlikely that today's networks will stand up to the demands of 2020.

As a potential solution, even software-defined networking only begins to address some of these needs in the cloud data center. However, it's a strong start in the right direction.

Hyper-connectivity will threaten the individual in more ways than currently acknowledged. We need to think about how to protect against inevitable threats to the system. Power grids will be more efficient, and consumer interactions such as net metering will drive the integration of operational technology (OT) and information technology (IT).

However, the integration of OT and IT opens the door for external threats. Cars will be able to drive themselves, controlled by applications. Having your networked garage recognize whether or not your car is in the garage is useful, but it also creates opportunity for hackers. Businesses and homes will become increasingly targeted and hackable. It's important that the networking industry works now to ensure security within these networks of tomorrow.

The Internet of Things represents huge changes underway across all industries. As sensors become more common in the home and at work, it's easy to imagine a future overtaken by connected objects and devices and a network that is unprepared for it. We need to think proactively about our role as an industry and start solving today the problems of a not-too-distant tomorrow.


agile networking

PG, Can you elaborate on how software-defined networking only begins to address the IoT needs for agile networking? What else needs to happen?

Re: agile networking
First, it is about the scale of the network. IoT is about billions while SDN today is focused on millions. One of the primary market drivers of SDN was the need to increase the scale by a few orders of magnitude, i.e., from being able to handle thousands of users to millions and the automation and agility that would be required to operate efficiently at that scale. Second, it is about administrative domains. While SDN is currently addressing automation and self-service through abstraction and orchestration, it assumes a fair degree of sophistication by its user. IoT will be about self-service for the common man. This requires a much higher level of abstraction and simplicity. And then there are other areas of IoT such the need for virtual distributed datacenters closer to the user because of real-time needs. SDN today is mostly about automation and agility for a (few) large data center(s).
Re: agile networking

I think you have hit the nail on the head -- without the Cloud, data analytics would be very difficult and without data analytics it is difficult to create value that drives the functionality of a device higher. And without a network capable of handling the data needs, value created cannot be implemented.

It is this interlinked nature of the IoT that makes it extremely interesting. Whether, networks are to become vastly more centralized or decentralized depends a lot on the type of products that will constitute the majority of those 50 billion connections. Either way, the network is a good place to understand the changes that are on their way. 

SDN is important in the network to enable network orchestration, etc. I feel that SDN is also important on the business side, as it enables the kind of long term relationships between data service providers and device producers that will be needed.