A cross-layer packet scheduling and antenna selection scheme for multi-service MIMO systems

In this paper, we propose a cross-layer packet scheduling scheme for multi-service MIMO system which adopts a unified method to select the set of users to transmit to and the antennas over which users send their packets. Our proposed scheduling scheme makes a difference between RT(Real Time) and non-RT(non-Real Time) services. It defines the utility functions of RT services according to a Modified Delay PF algorithm (MDPF), which could guarantee the QoS of RT services while combining the predominance of PF algorithm. And PF mapping is used for calculate the utility functions of non-RT services considering the tradeoff between throughput and fairness. Through a low-complexity antenna selection algorithm, set of users and antennas are determined at the same time according to the utility functions of every user. Simulation results reveal the proposed scheme provides better QoS guarantee to RT services while not badly deteriorating the throughput of system and fairness among users compared to the conventional PF scheduling.