Are there IoT applications that don’t need to implement security measures?

Any device that connects to the internet may be vulnerable to local or remote attacks. Attackers can target almost any connected device to try and steal manufacturer intellectual property stored in the system, gain access to user data, or even maliciously manipulate the system to compromise users or attack third parties online.

As demonstrated by the major distributed denial of service (DDoS) attack late last year, labeled by experts as the largest of its kind in history, even seemingly harmless products such as home digital video recorders (DVRs) can be maliciously infected and used as “botnets” to halt operations for third-party entities. The attack last year affected services such as Twitter and PayPal, but similar attacks could potentially target large smart infrastructure technologies such as electric grid systems. According to a 2016 study conducted by Kapersky Lab, a single DDoS attack can cost an organization more than $1.6 million to resolve.

Should IoT security features primarily focus on Wi-Fi and internet-level encryption of packets sent over the air?

Figure 1: Local network security features in action

Figure 2: CC3220 security features

Do you need to use a high-end microprocessor unit (MPU) or dedicated secure element to effectively protect your products against potential risks?

While increasingly lean IoT system resources often present design challenges, you can still strive to target more robust security in MCU-based, bill of materials (BOM)-optimized systems. The first steps are to identify which system assets are at risk, where the potential exposure points exist and what threats you anticipate will put the system at risk. From there, you work to choose components that offer a wide range of integrated hardware-based security features, while offloading the host MCU.

• Offload the processing of the device’s security functionality (including secure file system management) to the network processor and hardware cryptographic engines. This enables the million instructions per second (MIPS) and application MCU memory to be solely dedicated to the host application.
• Help reduce the firmware and network-management vulnerability for application-related risks by physically partitioning the programmable application MCU from the read-only memory (ROM)-based network processor