针对正交频分复用(orthogonal frequency division multiplexing,OFDM)系统多用户中继场景中混合业务对时延的不同需求,提出一种基于合理反馈机制下的降低时延、保证混合业务服务质量的调度算法(mixed services with feedback scheduling,MSFS)。MSFS算法针对用户实时业务及非实时业务的不同需求分别采用不同的调度准则及反馈方式,以达到优化系统性能的目的。针对实时业务,增大其调度优先级以减少等待时延,并在其等待时延接近时延门限时,使其调度优先级迅速增加,以避免丢包;针对非实时业务,加入时延公平因子,使得该类用户业务在时延公平性方面得到改善,同时设置适当的反馈门限,在不影响业务服务质量需求的同时达到降低系统反馈量的目的。仿真结果表明,该算法在保证系统吞吐量和用户公平性的同时,不仅能够有效降低实时用户的时延,而且能够较大程度地减少系统反馈量。
In this paper, we have proposed a multi-relay selection and power allocation scheme for two-way relay network which aims to maximize the sum-rate of two-way relay system. First, to prolong network lifetime, a multi-relay selection strategy is proposed in which both channel state information （CSI） and remaining energy （RE） are considered. Next, a multi-relay power allocation algorithm based on convex optimization （MRPA-CO） is presented. To reduce the computational complexity, it can be divided into two steps： terminal nodes power allocation （TNPA） and relay nodes power allocation （RNPA）. Simulation results indicate that the proposed relay selection strategy can significantly prolong network lifetime compared to other relay selection strategies which consider CSI only, and the MRPA-CO algorithm has great advantage over equal power allocation （EPA） on sum-rate in two-way relay network.
