In among the more doubtful IoT devices—the anti-slurping noodle fork, a treat dispenser for dogs, remote connected floss dispensers—there's actually a lot of real-world business applications benefitting from the technology:
Houston-based Accessible Home Health Care use
to allow senior patients to communicate remotely with hospital staff;
Swedish consultancy Tyrens AB
throughout its headquarters to monitor and control all aspects of the building environment in real-time;
Bell, Canada’s largest communications company, have developed
to help meet food safety regulations and improve product traceability.
Between consumer and business applications, there are seven billion devices in use, with that number set to increase to 10 billon by the end of 2020, equating to 62 million devices added each month for the next two years.
One of the main drivers is digital transformation programs that require data to be collected from the edge of the business—typically operational locations or customer touch points—to improve organizational knowledge and rapid decision making. As a result, there’s now a huge range of IoT devices available, spanning medical devices, industrial sensors, cameras, wearables, home appliances and a lot more besides.
But, as the use of IoT devices increases, so too does the ask of network infrastructure: downtime or performance limitations that delay transmission of business-critical data will have an immediate and costly impact on business operations.
Most traditional networks aren’t built to cope with the large number of distributed devices nor the varied and much bigger data volumes associated with IoT deployments. As a result, organizations that plan on increasing their IoT usage need to review and, where necessary, change their end-to-end network to cope with the need to:
Connect a large number of heterogeneous devices at low cost;
Add more devices at short notice and cope with the resultant increase in traffic flows;
Ensure rapid, reliable and secure data collection and transfer to centralized or cloud-based business apps;
Monitor and manage traffic at the device level; and
Prioritize traffic flows to ensure data is received when needed and transmission isn’t compromised by other applications.
One of the biggest challenges is providing low-cost connections for what could be hundreds or even thousands of IoT devices, some of which will use batteries and have no AC power supply.
A wired connection is often the most reliable but is practical only where wired networks are prevalent. Most current IoT devices connect using a Wireless Personal Network (WPAN), such as Bluetooth or ZigBee, but these have a limited range and occasional reliability problems. Wi-Fi is usually a widely available alternative but requires a lot of power and might already be operating at capacity in support of a mobile workforce.
Low-power Wide Area Networks (LPWAN) are relatively new—Sigfox, Lora, and NB-IoT are currently the main technologies—and account for about 10% of current IoT deployments. But adoption is expected to grow rapidly since it promises high battery life and a maximum communication range of over 20 kilometers. For the same reasons, it’s viewed as a better option than cellular technologies like 4G.
While a lot of IoT networking discussion understandably focusses on the edge, there are also impacts on core networking. Data gathered at the edge (and there will be a lot more of it) will, in many cases, need to be moved back to central data systems or distributed cloud apps for analysis and decision making. This means it might not be possible to rely on the same architectural design as used for traditional data environments.
Moreover, ruggedized networking equipment might be needed to support outdoor connectivity or operation in harsh environments.
A large number of widely distributed IoT devices introduces management and control challenges. Network administrators must be able to spot usage spikes that suggest additional resources need to be provisioned; detect unusual behavior indicative of a possible cyberattack; and easily add, configure or remove devices. Fortunately, products like Edgewater’s EdgeView Service Control Centre and Cisco DNA Center help make those tasks a lot easier.
The administrative burden can be reduced even more by designing the network to automatically detect devices as they are connected and immediately classify them in categories with pre-assigned QoS and access rights.
A 2017 survey found 48 percent of firms had experienced at least one IoT security breach, and by 2020 it's estimated a quarter of cyberattacks will target IoT devices. The DDoS attack that downed Dyn (and much of the US internet traffic) in 2016 was caused by IoT devices infected by the Mirai botnet, and check out this example that illustrates how criminals are now targeting relatively innocuous IoT devices as a way of breaching the corporate network and the data it holds.
Most IoT devices have little or no integrated security protection, which means a multi-layer security approach is essential. Besides the usual network security tools and techniques, every device should be identified, tracked, authenticated, and authorized.
The Gentek team have been working closely with our partners—Cisco, Edgewater, Grandstream, NETGEAR and SMARTRG—to ensure there’s a range of IoT solutions to help meet your customer’s requirements. Contact us today to find out how we can help.