As the Internet of Things becomes more widespread, the amount of data it produces increases exponentially. As that happens, it becomes both cumbersome and inefficient to rely on centralized servers and platforms. Instead, computing increasingly must be done at the edge. However, this comes with its own set of cybersecurity concerns and challenges.

The Internet of Things (IoT) is seeing explosive growth. According to a forecast from International Data Corporation, there will be approximately 41.6 billion connected IoT devices by 2025. Between them, they will generate approximately 79.4 zettabytes of data.

For context, a single zettabyte is so large that it would take approximately one million supercomputers to store it

That data will contain a wealth of insights on just about everything. Information that can be used to make smart cities smarter, safer, and more efficient. Details on everything from customer preferences and habits to bottlenecks in office workflows.

Suffice it to say, that’s not data that can simply be discarded. Rather, for the IoT to truly move forward, it’s imperative that the data it generates be analyzed, collected, and stored. In order to make this feasible, we need to transition to edge computing.

It’s functionally what it sounds like. Rather than transmitting data to a centralized server to be analyzed, analysis and generation are done as close to the data’s point of origin as possible. Only the insights are transmitted and stored. 

Unfortunately, just as traditional computing and networking infrastructure cannot support edge computing, so too is the traditional model of cybersecurity insufficient. Instead, edge computing requires a different approach, one which acknowledges that the traditional security perimeter is dead and gone. This requires a few things, according to Data Center Knowledge:

  • Shifting away from password-based security. Security experts have made no secret of the fact that passwords are obsolete. Poor password strength is also one of the chief cybersecurity issues associated with IoT. While it’s currently unclear what authentication systems will replace passwords, they need to die for us to move forward.
  • Expand network segmentation. Edge computing devices don’t always need to be directly connected to enterprise networks. In many cases, they shouldn’t be. Air gapping ensures that even if an edge device is compromised, it won’t put other systems and data at risk.
  • Connect only when necessary. Some edge devices might require always-on connectivity. Most don’t. By reducing the amount of time a device is connected to the Internet, you reduce the chances that it will be compromised.
  • Automate. The sheer number of edge devices is such that no human security professional should be expected to keep up with them. Artificial intelligence-based cybersecurity solutions can actively monitor the edge, notifying human actors only when something suspicious occurs. 
  • Better lifecycle management. Security patches and updates can no longer be put off. For edge devices, they must be applied immediately.
  • Prioritization of physical security. Securing a server room or physical desktop is easy. Securing a sensor rig or IoT appliance? Not so much. There must be some means of monitoring and safeguarding IoT hardware in addition to software, lest a criminal simply makes off with a device for the data it contains. 

Edge computing and the IoT represent a completely new frontier from a cybersecurity standpoint. However, the challenges that come with them are not insurmountable. At the end of the day, they simply require a slight change in perspective.