论文部分内容阅读
Modified largest weighted delay first(M-LWDF)is a typical packet scheduling algorithm for supporting hybrid real-time services over wireless networks.However,so far,there is little literature available regarding the theoretic analysis of M-LWDF fairness.This paper gives a theoretic analysis of M-LWDF fairness,which shows that M-LWDF fairness is related to channel condition,packet’s arrival process and the ratio of quality of service(QoS)requirements of different service queues.Given service QoS requirements and other parameters related to channel model and packet’s arrival process,the fairness is merely related to the ratio of the number of users in the service queues.Based on the analysis,an enhanced M-LWDF algorithm(EM-LWDF)is proposed and demonstrated in this paper.EM-LWDF is strictly designed in light of the fairness criteria of QoS requirements,so its fairness is almost not related to the ratio of the number of users in the service queues,and the theoretical value of fairness index is equal to 1.Simulation results validate the theoretic analysis and show the effectiveness of EM-LWDF in improving fairness.
Modified largest weighted delay first (M-LWDF) is a typical packet scheduling algorithm for supporting hybrid real-time services over wireless networks. Even, so far, there is little literature available regarding the theoretic analysis of M-LWDF fairness. This paper gives a theoretic analysis of M-LWDF fairness, which shows that M-LWDF fairness is related to channel condition, packet’s arrival process and the ratio of quality of service (QoS) requirements of different service queues.Given service QoS requirements and other parameters related to channel model and packet’s arrival process, the fairness is merely related to the ratio of the number of users in the service queues. Based on the analysis, an enhanced M-LWDF algorithm (EM-LWDF) is proposed and demonstrated in this paper. -LWDF is strictly designed in light of the fairness criteria of QoS requirements, so its fairness is almost not related to the ratio of the number of users in the service queues, and the theoretical value of fairness inde x is equal to 1. Simulation results validate the theoretic analysis and show the effectiveness of EM-LWDF in improving fairness.