Analysing and Learning Low-Latency Network Coding Schemes

Publikation: Beitrag in Buch/Konferenzbericht/Sammelband/GutachtenBeitrag in KonferenzbandBeigetragenBegutachtung

Abstract

Forward Error Correction (FEC) has become an integral part of communication technology to address expected losses during transfer of data. Among various layer three block codes, Random Linear Network Coding (RLNC) has emerged as an adaptable, powerful approach. However, its most straightforward implementation, Full Vector Coding (FVC), introduces too much delay for widespread adoption. Handcrafted schemes were introduced to optimise coding delay, while keeping resilience reasonably high. These works have resulted in tailRLNC and PACE. For the first time we analyse their respective behaviours in a fair and comparable manner, as non-recovered packets were statistically ignored previously. We then introduce an environment that uses a consistent, unbiased simulator and interface it with a Deep Reinforcement Learning (DRL) agent. This is the first time RLNC is joined with DRL. Our deep Q-network (DQN) based agent effectively uses an optimisation loop and utilises a customisable, expressive and extendable parametric loss function to learn a protocol. We demonstrate our agent recovers hand-tailored schemes and achieves state of the art.

Details

OriginalspracheEnglisch
Titel2023 19th International Conference on Wireless and Mobile Computing, Networking and Communications, WiMob 2023
Herausgeber (Verlag)IEEE Computer Society
Seiten175-180
Seitenumfang6
ISBN (elektronisch)979-8-3503-3667-2
PublikationsstatusVeröffentlicht - 2023
Peer-Review-StatusJa

Publikationsreihe

ReiheInternational Conference on Wireless Communications, Networking and Mobile Computing
Band2023-June
ISSN2161-9646

Konferenz

Titel19th IEEE International Conference on Wireless and Mobile Computing, Networking and Communications
KurztitelWiMob 2023
Veranstaltungsnummer19
Dauer21 - 23 Juni 2023
Webseite
OrtPolytechnique Montreal
StadtMontreal
LandKanada

Externe IDs

ORCID /0000-0001-8469-9573/work/161891136

Schlagworte

Schlagwörter

  • Network Coding, Reinforcement Learning