Chapter 3 - Lightweight code self-verification using return-oriented programming in resilient IoT

Abstract

The internet of things (IoT) represents the widespread deployment of devices with sensors in the physical world that connect with one another via the Internet. Protecting these IoT nodes from unapproved changes is essential to their unhindered functioning. Any possible leeway in the code creates vulnerability and allows an adversary to intervene and mount an attack on a victim’s system. This chapter establishes novel methods for protecting IoT devices from being compromised. We do this by turning offensive intrusion techniques into defensive weapons using return-oriented programming (ROP) gadgets. We developed a prototype IoT testbed at our laboratory consisting of IoT nodes, gateways, and a server. We also developed an approach using ROP to prevent code from being tampered with. Our approach is a two-layer code-protection mechanism using ROP and code verification via checksumming. Both layers guard each other and protect sensitive parts of a program. This mechanism strengthens the defense shield of IoT nodes. The experimental results show that the performance overhead to mount this type of security is negligible. It is infeasible that an adversary can bypass both layers of security at the same time. If an adversary tries to bypass the security shield then one of the layers alerts the administrator, therefore increasing the cost of exploitation.