TRAQR: Trust aware End-to-End QoS routing in multi-domain SDN using Blockchain

Abstract

Recent advances in today’s Internet introduced huge demand for End-to-End Quality of Service (E2E QoS) assurances for delivering different types of traffic (video, live streaming, and telemedicine), originating from a wide spectrum of applications. Such applications stretch across multiple administrative domains and require fair and robust coordination for E2E QoS provisioning. While the selection of trusted domains is of paramount importance for the successful realization of E2E QoS, the issue of domain’s trust while finding E2E QoS paths has not been actively explored in the existing works. Trust awareness reduces the risk of violating QoS and improves the confidence of operating QoS across multiple domains. However, identifying trusted domains, finding the source of trust, and quantifying trust are the major challenges for enabling trust aware E2E QoS routing. Auditability of QoS implementation results at each domain forms a vital source for verifying QoS compliance and subsequently evaluating the trust of the domain. However, securely obtaining logs from the domains and identifying the accountability of domains for E2E QoS, without relying on a third party introduces significant difficulty when evaluating domain trust with respect to QoS compliance. This work proposes Trust aware E2E QoS Routing (TRAQR) framework for provisioning E2E QoS with trust and verifying the QoS compliance in Multi-Domain Software Defined Networks (MD-SDN). TRAQR efficiently leverages the tamper-proof and decentralized properties of Blockchain to store and exchange different kinds of trust information required to provision and verify E2E QoS compliance of the domains. This work developed a proof of concept implementation of the TRAQR in an MD-SDN environment emulated using Mininet, RYU SDN controller, Ethereum Blockchain and Smart Contracts to demonstrate the performance and benefits of TRAQR. Further, the simulations showed that the domain trust parameter introduced significant improvement in QoS compliance, even in the presence of untrusted domains.