Comparison of QoS optimisation techniques in adaptive smart device peer-to-peer overlay networks. Issue 6 (2nd November 2021)

• Record Type: Journal Article
• Authors: Maiti, Ananda; Maxwell, Andrew; Kist, Alexander
• Is Part Of: International journal of parallel, emergent and distributed systems. Volume 36:Issue 6(2021)
• Page Start: 668
• Page End: 686
• Publication Date: 2021-11-02
• Subjects: Distributed computing -- smart devices -- peer-to-peer network -- assignment problem -- remote instrumentation -- data clustering; Parallel computers -- Periodicals; Electronic data processing -- Distributed processing -- Periodicals; Computer algorithms -- Periodicals; 004.35
• DOI: 10.1080/17445760.2018.1441840
• Languages: English
• ISSNs: 1744-5760

Abstract

Abstract: Smart device networks form the backbone of many Internet of Things (IoT) applications. The performance of these systems depended on many factors. This includes their ability to connect to each other and network latency. In this paper, two algorithms are discussed that build a Peer-to-peer (P2P) overlay network. The focus is on real-time end-to-end remote instrumentation for IoT applications. P2P overlay networks are used for various purposes from content distribution through to media sharing. In contrast to traditional P2P systems, IoT systems can be highly sensitive to network latency and require low bandwidth. The main contributions of this paper are two algorithms to create overlay networks that assign peer nodes to super-peer nodes. The network contains a large number of smart devices and the key aim is to minimize system wide network latency. Other constraints regarding the nodes ability include the ability to become a super-peer node such as connectivity and limited processing capacities of the smart devices are also considered in context of a P2P remote instrumentation system. Both algorithms create a 3-hop network where the latency is lowered as much as possible with the given set of nodes. The algorithms have been validated through simulation.