A Revenue-Based Low-Delay and Efficient Downlink Scheduling Algorithm in OFDMA Systems

In this paper, we proposed a revenue-based low-delay and efficient (R-LDE) algorithm for downlink packet scheduling in OFDMA systems. In actual operation of the network, telecommunication operators always expect the most revenue. Therefore, a simple revenue model is introduced and integrated into our proposed packet scheduling algorithm. The significances of R-LDE algorithm are threefold: firstly, it can guarantee the delay requirement of certain traffic; secondly, it can keep the efficiency of the system; thirdly, it considers the revenue of the telecommunication operators. The performance of R-LDE is evaluated in terms of packet delay, throughput and revenue with actual traffic models. It is then compared with conventional multi-carrier proportional fairness (MC-PF) algorithm and max SNR algorithm with mixed traffics such as FTP and video traffics. Simulation results show that the R-LDE algorithm can achieve the most scheduling revenue and performs much better than the algorithms stated before in terms of packet delay and throughput with video traffic at the cost of slight deterioration of FTP traffic performance.