Mobile cloud computing is one of the facets of cloud based systems, whereby mobile nodes obtain services from a global remote cloud platform in a more ecient way with respect to local service execution. Unfortunately, recent forecasts on cellular bandwidth (that is the key enabler for this paradigm) pose signicant challenges to the practical applicability of this approach. In this paper, we explore a complementary mobile cloud computing solution, where mobile nodes can also rely on other nodes in the vicinity that could provide the sought service. These nodes are contacted via direct communication based on WiFi or Bluetooth, which therefore ooads trac from the cellular network. In the proposed system, mobile nodes decide dynamically whether to access global or local cloud services based on the availability of the latter in their vicinity, and the load on the cellular network. Simulation results show that this solution provides lower average service provision times with respect to an alternative based exclusively on a remote cloud. As a side eect, such a system avoids cellular congestion and possible saturation, even in case of signicant load.
Offloading service provisioning on mobile devices in mobile cloud computing environments
Marco Conti;Davide Mascitti;Andrea Passarella
2015
Abstract
Mobile cloud computing is one of the facets of cloud based systems, whereby mobile nodes obtain services from a global remote cloud platform in a more ecient way with respect to local service execution. Unfortunately, recent forecasts on cellular bandwidth (that is the key enabler for this paradigm) pose signicant challenges to the practical applicability of this approach. In this paper, we explore a complementary mobile cloud computing solution, where mobile nodes can also rely on other nodes in the vicinity that could provide the sought service. These nodes are contacted via direct communication based on WiFi or Bluetooth, which therefore ooads trac from the cellular network. In the proposed system, mobile nodes decide dynamically whether to access global or local cloud services based on the availability of the latter in their vicinity, and the load on the cellular network. Simulation results show that this solution provides lower average service provision times with respect to an alternative based exclusively on a remote cloud. As a side eect, such a system avoids cellular congestion and possible saturation, even in case of signicant load.I documenti in IRIS sono protetti da copyright e tutti i diritti sono riservati, salvo diversa indicazione.