Friedrich-Alexander-Universität UnivisSuche FAU-Logo
Techn. Fakultät Willkommen am Institut für Informatik FAU-Logo
Logo I4
Computer Science 4
Operating Systems
Distributed Systems
Research Projects
Student Projects
Phone 09131/85-28820
Fax 09131/85-28732

University of Erlangen
Computer Science 4
Dr. Frank Bellosa
Martensstraße 1
91058 Erlangen

Room 0.037
Department of Computer Science  > Computer Science 4  > Frank Bellosa  > Student Projects
Application-Specific Power Management for Wireless Networks
Matthias Faerber
Advisor: Andreas Weißel, Dr.-Ing. F. Bellosa
Registered as Studienarbeit SA-I4-2004-02 , January 19 2004
[Abstract] [Full Paper (pdf) , 264 kB]

In recent years computer networks have gained importance in many fields. After the success of wired networks and the internet, radio based wireless networks are spreading. In many cases wireless networks are used when it is impossible to connect the computer to a network using a cable or it is just uncomfortable. This not only applies to network cables but also to power cables and so most devices are battery powered.

Wireless network adapters that are using the IEEE 802.11b standard are already equipped with a power management mode. In this power management mode the card continously changes between a sleep and an awake state. The sleeping interval can be altered by setting the so called beacon interval.

This work examines how network applications can be identified only with information available at the network layer. It is shown that statistical information derived from the linux kernel can be used to characterize and identify applications. For a set of applications such profiles are developed and explained.

Based on these profiles an algorithm is provided that is capable of detecting applications during their runtime and adjust the beacon interval of the IEEE Standard. Thus the static behavior of the power management mode it changed into a dynamic one. The algorithm considers both application performance and energy conumption and sets the power management to an user-defined trade-off between both.

The results of the characterization algorithm are evaluated in two tests. First offline tests are used to analyze the detection rates for one application alone. Second, online tests are used to analyze the delay between the start of an application and its detection and the impact of false characterizations.

About our Server | Mail to Webmaster
Last modification: 2/10/2005