Trace-Based Evaluation of Rate Adaptation Schemes in Vehicular Environments

There has been a variety of rate adaptation solutions proposed for both indoor and mobile scenarios. However, dynamic channel changing conditions (e.g., temporal channel variation due to unpredictable traffic pattern) make it virtually impossible to guarantee the same evaluation environment for all these schemes. Moreover, developing these schemes exhaustively on actual hardware can be gruesomely long. In this work, we propose an integrated framework which utilizes empirical data gathered from the vehicular testbed to objectively compare different rate adaptation schemes for Vehicular Ad-hoc Networks (VANETs). Using this framework, we implemented some well-known adaptation schemes and evaluated their performance. The main contribution of this paper is a methodology to compare rate adaptation schemes in an environment which is both realistic and repeatable. In addition, our results shed new light on impact of various different environment and channel factors on the performance of different schemes.