Opportunistic Scheduling with Statistical Fairness Guarantee in Wireless Networks

In this paper, we present statistical fairness opportunistic scheduling (SFOS), a wireless scheduling algorithm with the objective of improving system throughput while providing statistical fairness guarantee. In particular, SFOS provides statistical fairness by using a virtual time variation, while improving system throughput by using our designed utility function which relates to the transmission rate. We develop a general analytical framework for SFOS, which shows that the fairness index is bounded by the utility. Further, we investigate the design rule of the utility function over Rayleigh fading channels and discrete transmission rates and present a reference design. Simulation results evaluate SFOS can significantly improve system throughput while providing statistical fairness guarantee.