A Battery-Aware Algorithm for Supporting Collaborative Applications

There are several significant ways in which the ubiquity of battery-powered devices impacts the field of collaborative computing. First, applications such as collaborative data gathering become possible. Also, existing applications that depend on collaborating devices to maintain the system infrastructure must be reconsidered. The problem is that collaborative applications often require end-user devices to perform background tasks that are not directly advantageous to the user. In this work, we seek to better understand how laptop users use their batteries and explore the cost associated with using a laptop in a common peer-to-peer network—Gnutella. Based upon our findings, we evaluate a battery-aware alternative to Gnutella’s ultrapeer selection algorithm. The most significant result of our study indicates that a large portion of laptop users can participate in system maintenance without sacrificing any of their battery. These results show great promise for existing collaborative applications as well as new applications, such as collaborative data gathering.