DIMSIM -- Device Integrity Monitoring through iSIM Applets and Distributed Ledger Technology

TL;DR

This paper presents a novel architecture using eUICC applets and distributed ledger technology to monitor device integrity in industrial environments, eliminating the need for additional hardware and ensuring trust among multiple stakeholders.

Contribution

It introduces a new distributed ledger-based system leveraging eUICC applets for device integrity monitoring without extra hardware in multi-stakeholder industrial settings.

Findings

Hashing data packets takes microseconds

Reading and writing to immutable database takes milliseconds

System provides efficient integrity monitoring with low overhead

Abstract

In the context of industrial environment, devices, such as robots and drones, are vulnerable to malicious activities such device tampering (e.g., hardware and software changes). The problem becomes even worse in a multi-stakeholder environment where multiple players contribute to an ecosystem. In such scenarios, particularly, when devices are deployed in remote settings, ensuring device integrity so that all stakeholders can trust them is challenging. Existing methods, often depend on additional hardware like the Trusted Platform Module (TPM) which may not be universally provided by all vendors. In this study, we introduce a distributed ledger technology-oriented architecture to monitor the remote devices' integrity using eUICC technology, a feature commonly found in industrial devices for cellular connectivity. We propose that using secure applets in eUICC, devices' integrity can be monitored and managed without installing any additional hardware. To this end, we present an end-to-end architecture to monitor device integrity thereby enabling all the stakeholders in the system to trust the devices. Additionally, we leverage the properties of immutable databases to provide robustness and efficiently to our model. In our primary evaluations, we measure the overhead caused by hashing our proposed data packets and performance of integrating an immutable database into our system. Our results show that performing hashing on our data packets takes order of microseconds, while reading and writing to an immutable database also requires only milliseconds.