Fine-Grained Access Control-Enabled Logging Method on ARM TrustZone

Abstract

Most applications for the Internet of Things operate on embedded systems. In particular, embedded devices intended for smart healthcare, smart homes, and smart cars generate logs containing sensitive user information. These logs must be protected from malicious users while also being accessible for legitimate users to utilize them for providing customized services. Unfortunately, the existing logging system only supporting one-to-one encryption based on a server-client model, so there are limitations in building a decentralized logging infrastructure for the hyper-connected era. In this paper, we propose a new secure logging method that supports one-to-many encryption and extends existing logging systems to a decentralized logging infrastructure. In the proposed method, log publishers are able to encrypt generated logs and distribute them to cloud storage in real time and can ensure that only authorized log subscribers access the logs. For one-to-many encryption, we apply a key-policy attribute-based encryption scheme which is suitable for logging systems. For reliability and efficiency of logs, we apply a key-derivation process that cooperates with one-way hash functions within a trusted execution environment. In a real time logging scenario, the proposed method is 93% faster and occupies 83% less storage space than when an original attribute-based encryption scheme is applied. In addition, performance-tunable parameters can optimize our method for various environments.

TB-Logger: Secure Vehicle Data Logging Method Using Trusted Execution Environment and Blockchain

“…In this system, a two-dimensional hash chain is used to ensure the integrity and continuity of logging data. Recently, an attribute-based encryption method was proposed to support fine-grained access control of EDR data [10]. In addition, D. Lu et al proposed TEE-based SQLite for embedded systems [28].…”

“…T datagen ≥ 2wT mac + T sign + T ws (10) If TB-Logger violates Eq. 10, data loss occurs as the basic data block recording time would take longer than the basic data block generation time.…”

“…To deal with the limitations of digital signature-based logging techniques, other researchers proposed trusted execution environment (TEE)-based methods that leveraged both a digital signature algorithm and a message authentication code (MAC) algorithm [9], [10]. These TEE-based works first define a basic data unit (e.g., a CAN message) and a TABLE 1.…”

A Searchable Encryption Scheme with Biometric Authentication and Authorization for Cloud Environments

“…For reflecting on the progress of our current research, we took into consideration the recent advances that have been accomplished in attribute-based encryption with keyword search [24][25][26][27][28] and attribute-based encryption with privacy protection and accountability [29][30][31][32][33][34][35].…”

Towards explainability in robotics: A performance analysis of a cloud accountability system

“…This opens the performance perspective (Ding et al, 2015), where the overhead incurred by intensive logging is non-negligible added to the concept of redundant logs. Finally, there is a legal perspective facing the problems with logs such as privacy (Miller, 2017) or cybersecurity to log files, which opens the use of BlockChain-based technologies (White et al, 2019) or encrypted decentralized logging infrastructures like the one proposed by Lee et al (2020), which are not explored in this research but leave the way open for further development and analysis.…”