Collaborative SLA and reputation-based trust management in cloud federations

Abstract

Abstract Industry and academia shift from the single cloud provider paradigm to cloud federations and alternative models, which orchestrate heterogeneous resources, such as Mobile Edge Computing and Fog Computing. In such complex environments, incorrect selection of deployment platform can lead to underwhelming application performance. This dictates the necessity of Service Level Agreements (SLA) and trust management services in order to enforce performance guarantees and enable customers to simultaneously evaluate their application’s performance and give performance indicators for future provider selection. In this paper, we propose a collaborative SLA and Reputation-based Trust Management (RTM) solution for federated cloud environment. The SLA service defines clearly the performance metrics and measures the actual performance of the deployed cloud applications. Based on the SLA, the RTM service of the collaborative solution leverages several technical and user’s experience metrics to compute the reliability of the cloud providers and the credibility of the customers. A proof of concept of the collaborative solution in a realistic federated environment is provided and validated. The corresponding experimental results demonstrate that it objectively computes the cloud providers’ reputation values under various scenarios.