Step 4 – identify threats

Within the IoT, there is a clear blending of the physical and electronic worlds. This results in relatively simplistic physical attacks that can be used to thwart a system's functionality. As an example, did the designers of the system include any integrity protections on the position of the cameras that provide data for parking enforcement correlation?

The amount of human involvement in the system also plays a significant factor in the types of attacks that could be used against a system. For example, if human parking enforcement officers aren't involved (that is, the system automatically issues citations for staying over the time limit), then the ability of the system that reads the license plates would have to be thoroughly examined. Could someone spoof a vehicle by simply swapping license plates, or deny the system the ability to read the plate by putting an obscuring layer on top of them?

The popular STRIDE model can be applied to IoT system deployments. Use well-known vulnerability repositories to better understand the environment, such as MITRE's common vulnerability and exposure database. Uncovering the unique threats to any particular IoT instantiation will be guided by the threat types (note that is also a good time to utilize attack/fault tree analysis for some implementations and deployments):

The application of the STRIDE model with the additional components that support the IoT can be seen in the following table: