Sway: Traffic-Aware QoS Routing in Software-Defined IoT

In this paper, we propose a traffic-aware quality-of-service (QoS) routing scheme in software-defined internet of things (SDIoT) network. The proposed scheme exploits the unique features of software-defined networking (SDN), such as flow-based nature, and network flexibility, in order to fulfill QoS requirements of each flow in the network. We consider two types of QoS routing strategies-delay-sensitive and loss-sensitive-for incoming packets from end-devices in the network. The former is devised to deal with delay-sensitive flows, and the latter deals with loss-sensitive flows, in order to maximize the overall network performance. We propose a greedy approach based on Yen's K-shortest paths algorithm to compute the optimal forwarding path, while considering the QoS requirements of each packet. Consequently, the SDN controller deploys adequate flow-rules at the forwarding devices in the network. Extensive simulation results show that the proposed scheme significantly reduces the end-to-end delay and the percentage of flows which violate QoS constraints compared to the benchmarks considered in the study. It is also observed that the proposed scheme adequately satisfies the QoS requirements for both type of flows in contrast to the existing schemes. In particular, with 2000 flows in the network, the proposed scheme achieves 13%, 14% and 15% (with AttMpls topology) and 38%, 37% and 39% (with Goodnet topology) reduction in QoS violated flows as compared to the existing LARAC, SPD, and MRC schemes, respectively.

Automated summary

This paper proposes a traffic-aware QoS routing scheme in SDN-based IoT networks, utilizing both delay-sensitive and loss-sensitive routing strategies to maximize network performance. By employing a greedy algorithm to compute optimal forwarding paths and deploying flow rules at forwarding devices, the proposed scheme significantly reduces end-to-end delay and QoS violations. Extensive simulations demonstrate the scheme's effectiveness in satisfying QoS requirements for different types of flows, outperforming existing schemes in terms of reducing QoS violations.