Centralized coded caching problem, in which a server with N distinct files, each with the size of F bits, serves K users, each equipped with a cache of capacity MF bits, is considered. The server is allowed to proactively cache contents into user terminals during the placement phase, without knowing the particular user requests. After the placement phase, each user requests one of the N files from the server, and all the users' requests are satisfied simultaneously by the server through an error-free shared link during the delivery phase. A novel coded caching algorithm is proposed, which is shown to achieve a smaller delivery rate compared to the existing coded caching schemes in the literature for a range of N and K values; particularly when the number of files is larger than the number of users in the system.
Improved delivery rate-cache capacity trade-off for centralized coded caching / Mohammadi Amiri, M.; Gunduz, D.. - (2017), pp. 26-30. (Intervento presentato al convegno 3rd International Symposium on Information Theory and Its Applications, ISITA 2016 tenutosi a Hyatt Regency Monterey Hotel, usa nel 2016).
Improved delivery rate-cache capacity trade-off for centralized coded caching
D. Gunduz
2017
Abstract
Centralized coded caching problem, in which a server with N distinct files, each with the size of F bits, serves K users, each equipped with a cache of capacity MF bits, is considered. The server is allowed to proactively cache contents into user terminals during the placement phase, without knowing the particular user requests. After the placement phase, each user requests one of the N files from the server, and all the users' requests are satisfied simultaneously by the server through an error-free shared link during the delivery phase. A novel coded caching algorithm is proposed, which is shown to achieve a smaller delivery rate compared to the existing coded caching schemes in the literature for a range of N and K values; particularly when the number of files is larger than the number of users in the system.Pubblicazioni consigliate
I metadati presenti in IRIS UNIMORE sono rilasciati con licenza Creative Commons CC0 1.0 Universal, mentre i file delle pubblicazioni sono rilasciati con licenza Attribuzione 4.0 Internazionale (CC BY 4.0), salvo diversa indicazione.
In caso di violazione di copyright, contattare Supporto Iris