Jasper is now part of Cisco

Learn More

Privacy Statement has been updated. Read statement

About Cisco

Jasper is now part of Cisco

Learn More

About Cisco

Privacy Statement has been updated. Read statement

IoT Security – Key considerations for protecting connected devices

preview-4
preview-5
preview-6
February 16, 2017

by

Sanjay Khatri

The heart of every IoT business is the connected devices, where managing security comes with a host of unique challenges. While any device can be at risk because it’s connected to the internet to send and receive data, the level of risk will vary depending on the context of how the device is used. Those scenarios of use will typically drive your security strategies from device to device.

preview

In part one of our series on IoT security, we introduced players in the IoT ecosystem, and how each contributes to an overall security solution for a connected services business. In this article, we’ll take a deep dive into security considerations at the device layer.

Best practices for IoT device security

To safeguard connected devices, you’ll want to consider the following layers of security. The extent to which you apply them will depend on the device and its associated risk.

  1. Device authentication – Make sure both the device hardware and software are authenticated when trying to access the network, and verified before receiving or transmitting Use cryptographically-generated digital signatures, as most devices will not have users waiting to input login credentials to access the network.
  2. User access controls – Build in mandatory or role-based access controls and password policies to limit user access to device components and applications, ensuring they only access the resources needed to do their jobs. If any component is compromised, access control ensures the intruder has as minimal access to other parts of the system as possible.
  3. Application access controls – Limit which applications can access the device and securely monitor data sent to the device using security protocols such as firewalls, Intrusion Prevention Systems (IPS) and/or deep packet inspection capability (a combination of both firewall and IPS), depending on the perceived level of risk.
  4. Device lifecycle management – Implement procedures that enable devices to get timely security patches and updates to firmware and software.
  5. Data encryption – Protect configuration bit streams with encryption, and implement secure cryptographic key storage and zeroization capabilities, which erase sensitive parameters to prevent their disclosure and access to data, when device tampering is detected.

Security is never one size fits all

For an IoT business, ideally you might implement the strictest security controls in every situation, but it’s cost-prohibitive when you’re talking about thousands or millions of devices. Typically, it’s worth the investment to ensure the highest security for devices with a high asset value – such as a connected car or jet engine – due to the complexity of the device, as well as the bigger picture impact if the device is hacked.

Furthermore, device data has different levels of sensitivity, some that warrant more robust protection than others. For example, the security risk of tampering with a device connected to a nuclear plant is considerably greater than a device connected for soil monitoring. For each device, it’s important to evaluate the level of risk you can reasonably expect, and invest security resources commensurate with that risk.

Application scenarios: How risk factors influence levels of security

So how do you evaluate security risk for a connected device? There are many complex considerations. For our discussion here, we’ll break it down in simple terms. Below we look at three different application scenarios that illustrate risk based on how a device is used and where it may fit in a larger IoT solution, and the security best practices that may need the most attention for each scenario.

Consumer applications – e.g., Connected Car, Smart Home

Hacking certain consumer devices could result in personal harm, so risk factors to consider may include:

  • Complexity of devicesConnected cars may use multiple devices or complex devices that manage multiple functions. Multiple types of connectivity in the car (e.g., Wi-Fi, Bluetooth, and automaker network) mean multiple entry points of exposure to risk.
  • Consumer safety – In both connected car and smart home security scenarios, the nature of data pushed and pulled from devices may be more sensitive such as personal and financial information, as well as proprietary company data.

Key Best Practice: As these devices may run multiple applications and open up layers of risk, pay close attention to user access controls to limit the exposure of devices and applications to outside risk.

Infrastructure applications – e.g., Smart City app, Smart Meter

While smart city apps for traffic routing or smart meters for utility monitoring may be low complexity devices, they are implemented on a massive scale affecting larger infrastructures. Some risk factors include:

  • Scale of impact – Wide scale implementation of devices impacts a large population of users and broader functionality. For example, one hack could knock out hundreds of thousands of routing devices, disrupting traffic flows citywide and potentially endangering drivers.
  • Potential for escalation – If someone hacks into a single smart meter, it’s annoying for the homeowner; however, if they hack into a cloud-based smart grid, it could result in a catastrophic data breach and leave millions of people without power.

Key Best Practices: These devices may be simple, but are in use a long time – so effective lifecycle management is important. Keep it simple with no user access and avoid running multiple applications on them to reduce exposure to risk. Make sure you can easily update devices with new firmware and security patches to minimize vulnerability.

Enterprise applications – e.g., Retail Point of Sale, Transportation & Logistics

These types of devices have a broad sphere of influence and exposure, so the risk factors grow even more complex:

  • Snowballing impact – Hacking that affects point-of-sale terminals often impacts millions of devices worldwide, affecting the devices and related systems, and more importantly, may compromise consumer financial data and merchant company data. It can also result in severe losses for the business in terms of consumer confidence and sales (think major breaches at Target and Home Depot in recent years).
  • Supply chain impact – Large-scale fleet management safety and security is impacted on many levels, from thousands of drivers and vehicles on the road along with their parent freight company, to valuable assets in transit, and the related cargo owners and customers.

Key Best Practices: Given the breadth and severity of risk with these devices, implementing robust data encryption capabilities is mission critical.

While IoT security is a complex puzzle, implementing the appropriate protocols at the device level can go a long way toward safeguarding your business and your customers, and help accelerate your IoT success.

In our next article in this IoT security series, we’ll look at IoT security at the network and data transport layers. Stay informed – subscribe to our blog! In case you missed it, check out our first blog of the IoT security series, “Mastering IoT security – It takes a village“.

Post a comment

Post a Comment:

Subscribe

By email

Follow Us