Exploiting stream scheduling in QUIC: Performance assessment over wireless connectivity scenariosShow others and affiliations
2024 (English)In: Ad hoc networks, ISSN 1570-8705, E-ISSN 1570-8713, Vol. 164, article id 103599Article in journal (Refereed) Published
Abstract [en]
The advent of wireless technologies has led to the development of novel services for end-users, with stringent needs and requirements. High availability, very high throughput, low latency, and reliability are all of them crucial performance parameters. To address these demands, emerging technologies, such as non-terrestrial networks or millimeter wave (mmWave), are being included in 5G and Beyond 5G (B5G) specifications. mmWave enables massive data transmissions, at the expense of a more hostile propagation, typical for high frequency bands. Consequently, the inherent instability of the physical channel significantly affects the upper layers of the protocol stack, resulting in congestion and data losses, which might strongly hinder the overall communication performance. These challenges can be addressed not only at the link layer, but at any affected layer. QUIC is a new transport protocol designed to reduce communications latency in many ways. Among other features, it enables the use of multiple streams to effectively manage data flows sent through its underlying UDP socket. This paper introduces an implementation of priority-based stream schedulers along with the design of a flexible interface. Exploiting the proposed approach, applications are able to set the required scheduling scheme, as well as the stream priorities. The feasibility of the proposed approach is validated through an extensive experiment campaign, which combines Docker containers, the ns-3 simulator and the Mahimahi framework, which is exploited to introduce realistic mmWave channel traces. The results evince that an appropriate stream scheduler can indeed yield lower delays for time-sensitive applications by up to 36% under unreliable conditions.
Place, publisher, year, edition, pages
Elsevier, 2024. Vol. 164, article id 103599
Keywords [en]
5G mobile communication systems, Data transfer, Wireless networks, End-users, Performance assessment, Priority-based schedulers, QUIC, Real-time application, Stream multiplexing, Stream Scheduling, Stringents, Wireless connectivities, Wireless technologies, Millimeter waves
National Category
Communication Systems Computer Sciences Telecommunications
Research subject
Computer Science
Identifiers
URN: urn:nbn:se:kau:diva-101424DOI: 10.1016/j.adhoc.2024.103599ISI: 001295473300001Scopus ID: 2-s2.0-85200995962OAI: oai:DiVA.org:kau-101424DiVA, id: diva2:1891898
2024-08-232024-08-232026-02-12Bibliographically approved