Industry discussions about the Internet of Things (IoT) usually quote big numbers -- e.g., 30 billion connected devices by 2022 and a global IoT market valued at $14.4 trillion. For service providers, there is an urgent need to scale up, meet those new network requirements, and capture their share of the IoT market opportunity.
But big-scale strategies alone don't work when applied to IoT security. Mobile network operators in the IoT space, such as Telenor Connexion, have already identified that a very small number of misbehaving devices (less than 0.01 percent of the base) can create a signaling storm, resulting in significant congestion that can cascade to other networks. Hackers deploy the same tactics because, unlike service providers, hacker economics do not depend on large-scale success. Malicious actors only need a few successful infections to be economically viable or to disrupt critical service.
Hacker infrastructure investment and cost barriers are quite low. Bad actors use inexpensive, mass-scale techniques (e.g., email, automated bots, off-the-shelf exploit kits) to canvass large quantities of networked devices and locate vulnerable targets, but only need to infect a small number of devices or network elements (or even just one) in the right location or network to be economically viable or launch a successful network attack.
Finding those relatively "few" infected devices, and the malicious traffic associated with them, needs to be the primary objective of service provider network security strategies.
Hunting for "the few" may seem counterintuitive to service providers, where high performance and throughput are critical criteria in infrastructure equipment decisions. But a successful security posture for billions of IoT devices requires just that -- accurate and rapid identification of that small percentage of the total traffic attempting to infect devices, and of those devices already infected. Higher performance, without a truly effective threat prevention approach, simply gives malware a high-speed "free ride." And malware doesn't need a high "adoption rate" to create damage.
In a published example, researchers demonstrated that malware infection rates of less than 0.1 percent of the population focused on specific 911 centers can severely impair the availability of critical emergency services. In the research model, the malware caused infected devices to generate "false" calls to the 911 center. The added "false" calls, which require a longer response time, tie up call center resources and effectively make the service unavailable for other genuine emergency calls.
So how can service providers locate and stop the "few" when there will be billions of devices connecting to their networks -- most of which they have little control over? Endpoint protection alone (such as anti-malware software on cellphones) is insufficient -- most mobile subscribers simply won't use it, and it's not feasible for the majority of IoT devices.
A network-based prevention program is a more realistic solution. Service providers must be able to "see" and inspect their traffic to determine whether it is malicious and identify already-infected devices on their network to prevent hacker success. By thinking small -- that is, focusing on the minor percentage of traffic that is malicious, or identifying infected devices -- can service providers offer their subscribers and IoT providers a safe network free from infected devices and security breaches.
Terry Young is Senior Manager of Service Provider Product Marketing at Palo Alto Networks, where she is responsible for developing programming to communicate the business value of security for mobile network operators and other service providers.