Wi-Fi hotspots are one of the most promising scenarios for mobile computing. In this scenario, a very limiting factor is the scarcity of mobile-device energetic resources. Both hardware and software architectures of current devices are very ine cient from this standpoint, mainly the networking subsystem. This work analyzes a power-saving network architecture for the mobile-Internet access through Wi-Fi hotspots. Specifically, this solution supports any kind of best-e ort network applications, since it is application-independent. In this paper we derive a complete analytical model of the power-saving system when applied to mobile Web access. Furthermore, we use this model to compare our solution with a well-known approach, i.e., the Indirect-TCP. The comparison is performed by considering two performance gures: the energy saved in downloading a Web page and the related transfer-time. The results show that, in the average, our solution saves up to 78% of the energy. Furthermore, the power-saving system introduces an additional average transfer-time of 0.4 sec, and hence it does not signi cantly a ect the QoS perceived by the users. Finally, we assess the sensitiveness of the power-saving system with respect to Internet key parameters, such as the available throughput and the RTT.
Power-Saving in Wi-Fi Hotspots: an Analytical Study
Conti M;Gregori E;Passarella A
2003
Abstract
Wi-Fi hotspots are one of the most promising scenarios for mobile computing. In this scenario, a very limiting factor is the scarcity of mobile-device energetic resources. Both hardware and software architectures of current devices are very ine cient from this standpoint, mainly the networking subsystem. This work analyzes a power-saving network architecture for the mobile-Internet access through Wi-Fi hotspots. Specifically, this solution supports any kind of best-e ort network applications, since it is application-independent. In this paper we derive a complete analytical model of the power-saving system when applied to mobile Web access. Furthermore, we use this model to compare our solution with a well-known approach, i.e., the Indirect-TCP. The comparison is performed by considering two performance gures: the energy saved in downloading a Web page and the related transfer-time. The results show that, in the average, our solution saves up to 78% of the energy. Furthermore, the power-saving system introduces an additional average transfer-time of 0.4 sec, and hence it does not signi cantly a ect the QoS perceived by the users. Finally, we assess the sensitiveness of the power-saving system with respect to Internet key parameters, such as the available throughput and the RTT.I documenti in IRIS sono protetti da copyright e tutti i diritti sono riservati, salvo diversa indicazione.
