FAST: A fuzzy-based adaptive scheduling technique for IEEE 802.16 networks

Since the IEEE 802.16 first standard was proposed in 2004 to provide broadband wireless service, the standard has not only been widely studied, but also broadly commercialized. The current IEEE 802.16-2009 standard document specifies five Quality of Service classes. As is typical with most standards, IEEE 802.16 does not require the use of a specific scheduler. In this paper, we first evaluate the performance of four popular schedulers. By analyzing the results, we highlight that no single scheduler type performs the best in all traffic situations; however, we shown that there exist the most favorable scheduler type in each situation. Based on this rationale, our idea is to propose an adaptive scheduling schema where the scheduler is dynamically chosen based on the current traffic context, such as the number of flows of each Quality of Service class. We investigate this approach and evaluate its performance against existing static schemas. The results show that our approach presents some interesting performances in terms of throughput, delay, and packet loss ratio regarding state of art approaches.