Circumventing plateaux in cellular data offloading using adaptive content reinjection

Abstract : Mobile data offloading is a promising strategy to alleviate the burden on the cellular network by leveraging unused bandwidth across different wireless technologies. In this paper, we focus on hybrid offloading techniques involving both Wi-Fi and device-to-device communications. In this context, designing efficient solutions is challenging, as communication opportunities are, by nature, dependent on both individual mobility patterns and availability of the Wi-Fi infrastructure. We propose, design, and evaluate DROiD (Derivative Re-injection to Offload Data), an efficient strategy to finely control the distribution of popular content in urban scenarios. The idea is to use cellular resources as seldom as possible. To this end, DROiD injects copies using the cellular network only when needed: (i) in the beginning, when the Wi-Fi infrastructure is unable to trigger the dissemination, (ii) if the evolution of the dissemination is below some expected pace, and (iii) when the delivery delay is about to expire, in order to guarantee that all destinations receive the data. Our strategy is particularly efficient in highly dynamic scenarios, where sudden creation and dissolution of clusters of mobile nodes prevent proper content diffusion. We assess the performance of DROiD by simulating a traffic information service over two realistic large-scale vehicular datasets with several thousands of nodes. DROiD substantially outperforms other offloading strategies (opportunistic and AP-based), saving significant amount of cellular traffic even in the case of tight delivery delay constraints and energy limitations.
Liste complète des métadonnées
Contributeur : Marcelo Dias de Amorim <>
Soumis le : vendredi 24 mars 2017 - 15:03:25
Dernière modification le : mercredi 21 mars 2018 - 18:57:58




Filippo Rebecchi, Marcelo Dias de Amorim, Vania Conan. Circumventing plateaux in cellular data offloading using adaptive content reinjection. Computer Networks, Elsevier, 2016, 106, pp.49-63. 〈10.1016/j.comnet.2016.06.012〉. 〈hal-01495073〉



Consultations de la notice