Advertisement

For someone who works in cloud computing, the Internet of Things (IoT) is a tantalizing prospect, looming on the horizon. The promise of an interconnected world, driven by communications, data and analytics, has its attractions. And given the speed of technology development, the distance to that horizon is shrinking fast.

That’s why it behooves us to think about the challenges we face in making IoT an effective source of insights to improve our quality of life. Real-world proxies for IoT, such as cloud computing and the Internet, already face constraints that need to be addressed.

The good news is that these constraints – storage and processing, security and privacy, to name a few – are recognized by the computing field and work on multiple paths to solutions is underway.

Storage and Processing: A Cliff?

The notion of a “cloud” is reassuring, a gauzy realm where big data is tamed and everything is possible. Though cloud computing appears to offer an infinite amount of data storage and processing, each cloud is supported by finite physical resources in terms of hardware and energy. People in the cloud computing field are concerned that the IoT might represent a tipping point where big data overwhelms our ability to store and process it. The IEEE IoT portal and the IEEE Big Data Initiative portal both provide a wealth of resources on these rapidly evolving topics. 

I give this notion some credence because we already have an issue with the storage and processing of the vast amounts of unstructured data generated by various embedded systems. Embedded systems have dedicated functions that are integrated within larger systems and they have limited abilities for processing that data. Microprocessors, for example, generate huge amounts of data that isn’t immediately used but it may contain valuable insights. So storage and processing of unstructured data is already a concern and projections for big data generated by IoT suggest that we’ll need to adapt existing strategies and technologies and develop new ones to cope with, if not master, this challenge.

One familiar strategy is to cut the challenge down a size. If we start at the sensor level at the network’s edge, preliminary processing or winnowing might send only a manageable portion of incessant data streams for central storage and processing. Another strategy is logical partitioning (LPAR), which divides data into manageable portions to glean meaningful insights. Another concept being developed is inter-cloud computing. If a single problem exhausts the physical limitations of a single cloud service and its supporting infrastructure, separate clouds might handle parts of the problem, reducing the challenge to storage and processing.

Security and Privacy: Perennial Concerns

Another issue we face is user trust of the Internet, which enables the cloud and, eventually, IoT. Today, the lack of transparency in how user-generated data is collected and applied by search engines and advertisers, among others, creates a gray area. We can’t live without the Internet, but many if not most of us have no idea how our usage data is collected and how it is applied. Data breaches and the occasional peek behind the curtain into current practices have exacerbated uncertainty.

Though many firms scrub such data of any personally identifiable information before they use it, this is not a universal practice and users have a right to know what they’re getting into. Of course, some consumers are happy to trade their personal data for value in return, but that trade-off is not always clear to the user. This is another current issue that will only grow in importance as we enter the IoT age. Perhaps the market will favor the enterprises that provide transparent data policies as a commercial differentiator and I think we’re already seeing that strategy emerge.

Resources Available

Obviously, in a brief blog, I’ve only touched on a few salient issues relating to cloud computing, IoT and Big Data at a very high level. The challenges and solutions are much more complex than my broad brush makes them appear. If you’re a student or young professional interested in tackling these challenges, the upcoming IEEE Technology Time Machine conference, 20-21 October, in San Diego, will provide an opportunity to network with your peers and with professionals currently in the field. And a cloud computing track will be held in tandem with the annual International Conference on Consumer Electronics (ICCE) in Las Vegas, 8-11 January 2017.

Advertisement
Advertisement