Skip to Main content Skip to Navigation
Journal articles

Analytical Derivation and Optimization of a Hybrid Unicast-Broadcast Network for Linear Services

Abstract : This paper provides a theoretical analysis of the performance of an hybrid network composed of a broadband system and a broadcast system which cooperate to deliver linear services on portable and mobile devices. First, we propose a global system model for the hybrid network in which the coverage area of the broadcast system is adjusted with respect to the spatial distribution of the users. Then based on the number of users requesting the services, the model parameters are optimized to increase the capacity of the hybrid network while guaranteeing a given quality of service. By applying convex optimization techniques, we derive a closed-form expression of the optimal coverage radius of the broadcast system that maximizes the capacity of the hybrid network. Next, based on this theoretical model, we analyze the performances of the optimized hybrid network. The results show that, depending on the number of users requesting the service, significant improvements of the capacity and quality of service are achieved with the hybrid network. Finally, regarding the planning of a cooperative network, this paper can be used to determine the optimal planning parameters of the broadcasting system for the delivery of linear services.
Keywords : delivery
Complete list of metadatas

https://hal-univ-rennes1.archives-ouvertes.fr/hal-01436770
Contributor : Laurent Jonchère <>
Submitted on : Monday, January 16, 2017 - 4:50:36 PM
Last modification on : Monday, October 5, 2020 - 9:50:30 AM

Identifiers

Citation

Pape Abdoulaye Fam, Stéphane Paquelet, Matthieu Crussière, Jean-François Hélard, Pierre Brétillon. Analytical Derivation and Optimization of a Hybrid Unicast-Broadcast Network for Linear Services. IEEE Transactions on Broadcasting, Institute of Electrical and Electronics Engineers, 2016, 62 (4), pp.890--902. ⟨10.1109/TBC.2016.2593403⟩. ⟨hal-01436770⟩

Share

Metrics

Record views

2093