Analysis of adaptation strategies for mobile QoS-aware applications

Abstract

In mobile computing environments the availability of resources may vary significantly and unpredictably at application runtime, and thus may not satisfy a negotiated QoS level. We assume that in the presence of QoS fluctuations applications adapt their behavior to the prevailing operating conditions. We present five new adaptation strategies for application adaptation and study their performance by simulation experiments. The experiments show that one strategy performs best in most configurations in terms of result quality and computational efficiency. This strategy takes into account the ratio of utility value over resource demand as well as the system state.