Blockchain-Based Framework and Simulation Middleware for Service Level Agreement Compliance Assessment in the Context of IoT

Abstract

The concept of Service Level Agreement (SLA) is commonly employed to regulate the con tractual relationships between service providers and consumers. While several domains relate to SLA, this thesis focuses on SLA in the context of IoT. Regardless, SLA can be fragile and susceptible to violations and trust issues without a reliable trust mechanism in place. Recently, Blockchain has presented itself as an appealing alternative for mitigating trust issues related to centralised authorities and third parties. Following the success of Bitcoin, several blockchain platforms have emerged, such as Ethereum and Hyperledger Fabric, to enable conducting dis trusted processes in a non-repudiable manner. This thesis adopts Hyperledger Fabric as an underlying blockchain infrastructure and examines how Blockchain can be incorporated to serve distrusted processes related to a typical SLA life-cycle (SLA definition, monitoring, compliance assessment, penalty enforcement, and termination). First, it explores the literature of both tradi tional SLA practice and blockchain-based SLA studies. Accordingly, it proposes and evaluates an SLA representation and awareness approach within the Blockchain and demonstrates its benefits for SLA definition and negotiation purposes. Following, it experiments with the use of Blockchain for SLA monitoring, compliance assessment, and penalty enforcement in the context of IoT. Hyperledger Fabric employs a mechanism for preventing the double-spending problem, usually associated with monetary applications. However, this thesis demonstrates that the MVCC protocol does not align well with the high rate of transactions expected from the monitoring tool. Therefore, it proposes a set of design considerations for smart contracts to resolve these issues. Accordingly, it evaluates the performance of the proposed solution and reports a considerable improvement compared to naive approaches. Finally, this thesis contributes a middleware to close the gap between IoT simulated environments and real-world Blockchain platforms. Thus, it facilitates the usage of IoT simulators for Blockchain-based SLA purposes in terms of workload generation, metrics monitoring and benchmarking.