We investigate the problem of providing a video streaming service to mobile users in an heterogeneous cellular network composed of micro e-NodeBs ( μ eNBs) and macro e-NodeBs (MeNBs). More in detail, we target a cross-layer dynamic allocation of the bandwidth resources available over a set of μ eNBs and one MeNB, with the goal of reducing the delay per chunk experienced by users. After optimally formulating the problem of minimizing the chunk delay, we detail the Cross LayEr Video stReaming (CLEVER) algorithm, to practically tackle it. CLEVER makes allocation decisions on the basis of information retrieved from the application layer as well as from lower layers. Results, obtained over two representative case studies, show that CLEVER is able to limit the chunk delay, while also reducing the amount of bandwidth reserved for offloaded users on the MeNB, as well as the number of offloaded users. In addition, we show that CLEVER performs clearly better than two selected reference algorithms, while being very close to a best bound. Finally, we show that our solution is able to achieve high fairness indexes and good levels of Quality of Experience (QoE).
Colonnese, S., Cuomo, F., Chiaraviglio, L., Salvatore, V., Melodia, T., Rubin, I. (2018). CLEVER: a Cooperative and Cross-Layer Approach to Video Streaming in HetNets. IEEE TRANSACTIONS ON MOBILE COMPUTING, 17(7), 1497-1510 [10.1109/TMC.2017.2774298].
CLEVER: a Cooperative and Cross-Layer Approach to Video Streaming in HetNets
Luca Chiaraviglio;
2018-01-01
Abstract
We investigate the problem of providing a video streaming service to mobile users in an heterogeneous cellular network composed of micro e-NodeBs ( μ eNBs) and macro e-NodeBs (MeNBs). More in detail, we target a cross-layer dynamic allocation of the bandwidth resources available over a set of μ eNBs and one MeNB, with the goal of reducing the delay per chunk experienced by users. After optimally formulating the problem of minimizing the chunk delay, we detail the Cross LayEr Video stReaming (CLEVER) algorithm, to practically tackle it. CLEVER makes allocation decisions on the basis of information retrieved from the application layer as well as from lower layers. Results, obtained over two representative case studies, show that CLEVER is able to limit the chunk delay, while also reducing the amount of bandwidth reserved for offloaded users on the MeNB, as well as the number of offloaded users. In addition, we show that CLEVER performs clearly better than two selected reference algorithms, while being very close to a best bound. Finally, we show that our solution is able to achieve high fairness indexes and good levels of Quality of Experience (QoE).File | Dimensione | Formato | |
---|---|---|---|
clever.pdf
solo utenti autorizzati
Licenza:
Copyright dell'editore
Dimensione
1.45 MB
Formato
Adobe PDF
|
1.45 MB | Adobe PDF | Visualizza/Apri Richiedi una copia |
I documenti in IRIS sono protetti da copyright e tutti i diritti sono riservati, salvo diversa indicazione.