File Size Distributions and Caching for Offloading

Joonas Paakkonen, Prathapasinghe Dharmawansa, Ragnar Freij-Hollanti, Camilla Hollanti, Olav Tirkkonen

Tutkimustuotos: Artikkeli kirjassa/konferenssijulkaisussaConference contributionScientificvertaisarvioitu

2 Sitaatiot (Scopus)
164 Lataukset (Pure)

Abstrakti

We consider order statistics of products of random variables to understand the relationship between traffic offloading and storage space consumption in an edge caching network. Given a statistical model for the popularities and sizes of data objects in a caching network, we derive the order statistics describing the probability density function of the size of the object that accounts for, e.g., the largest amount of data traffic. We further study the asymptotic behavior of these distributions, and give analytical results for both the general case and the asymptotic case. The derived results can be used to estimate an upper bound on the cache size needed when one is interested in offloading a certain fraction of all traffic for file catalogues with changing content but known distributions of popularities and object sizes.

AlkuperäiskieliEnglanti
Otsikko2019 IEEE 20th International Workshop on Signal Processing Advances in Wireless Communications, SPAWC 2019
KustantajaIEEE
ISBN (elektroninen)978-1-5386-6528-2
ISBN (painettu)978-1-5386-6529-9
DOI - pysyväislinkit
TilaJulkaistu - 1 heinäk. 2019
OKM-julkaisutyyppiA4 Artikkeli konferenssijulkaisuussa
TapahtumaIEEE International Workshop on Signal Processing Advances in Wireless Communications - Cannes, Ranska
Kesto: 2 heinäk. 20195 heinäk. 2019
Konferenssinumero: 20

Julkaisusarja

Nimi IEEE International Workshop on Signal Processing Advances in Wireless Communications
ISSN (painettu)2325-3789
ISSN (elektroninen)1948-3252

Workshop

WorkshopIEEE International Workshop on Signal Processing Advances in Wireless Communications
LyhennettäSPAWC
Maa/AlueRanska
KaupunkiCannes
Ajanjakso02/07/201905/07/2019

Sormenjälki

Sukella tutkimusaiheisiin 'File Size Distributions and Caching for Offloading'. Ne muodostavat yhdessä ainutlaatuisen sormenjäljen.

Siteeraa tätä