COSMIC - Coding for Optimized Sending in Multimedia InterPlanetary Communication

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

Abstract

A popular technique for enhancing Content Delivery Networks (CDNs) is leveraging the opportunities provided by Peer to Peer (P2P) networks, of which the Interplanetary File System (IPFS) is a widely used one. While IPFS tends to perform well only in homogeneous networks, performance tends to quickly deteriorate when the channel parameters of individual peers differ, which is the case in real-world scenarios. A crucial technology for facilitating the flexible distribution of data requests is Random Linear Network Coding (RLNC). The RLNC technology is leveraged by the protocol Storing coded Packets In-advance For IPFS (SPIFI). Based on SPIFI, we design and implement the Coding for Optimized Sending in Multimedia InterPlanetary Communication (COSMIC) protocol, which uses an adaptive data request scheme to fit all imaginable network configurations. We compare the performance of the three protocols, IPFS, SPIFI, and COSMIC, in tailored and random test cases. To quantify performance, we use the transmission time and transmitted overhead as our metrics. Regarding the transmission time, COSMIC beats the other protocols by at least 3 0%. The transmitted overhead of COSMIC is slightly worse than SPIFI, but still beats IPFS by a wide margin. For comparison, we also calculate the minimal theoretically possible transmission time in a distribution network with random parameters as a lower bound.

Details

OriginalspracheEnglisch
Titel39th International Conference on Information Networking, ICOIN 2025
Herausgeber (Verlag)IEEE Computer Society
Seiten390-395
Seitenumfang6
ISBN (elektronisch)979-8-3315-0694-0
ISBN (Print)979-8-3315-0695-7
PublikationsstatusVeröffentlicht - 2025
Peer-Review-StatusJa

Konferenz

Titel39th International Conference on Information Networking
KurztitelICOIN 2025
Veranstaltungsnummer39
Dauer15 - 17 Januar 2025
Webseite
OrtKantary Hills Chiang Mai
StadtChiang Mai
LandThailand

Externe IDs

ORCID /0000-0001-8469-9573/work/185739510

Schlagworte

Schlagwörter

  • Distributed Storage, IPFS, Multimedia Content Distribution, Multisource Transmission, RLNC