基于吞吐量和延迟折衷的 LTE 下行调度算法研究

在LTE(long term evolution)无线网络下行链路系统中,调度算法是其无线资源管理的重要技术,其主要任务就是为无线用户传输的各种业务合理地分配无线资源.尤其是针对Non-GBR(non-guranteed bit rate)业务传输,可通过一个高效的实时调度算法来提高系统的吞吐量,且保证每个用户的服务质量,其在很大程度上决定了整个系统的性能.鉴于最大权重(maxweight)调度算法虽然使系统的吞吐量最佳,但是却导致单个用户的延迟较大的缺点,运用李雅普诺夫最优化的理论提出了一种实时调度算法,此算法不需要知道信道状态概率的先验知识,仅根据当前的信道状态和当前的队列积压来做实时的资源分配决定和传输决定.理论分析和仿真结果表明,该算法在吞吐量和延迟之间有一个很好的折衷,不但使网络吞吐量最优化,而且减小了单个用户的延迟.

A research on tradeoffs scheduling algorithm between throughput and delay in the downlink of LTE

In the downlink of LTE( long term evolution) wireless networks system, the scheduling algorithm is an important technology for the radio resource management, and it''s main task is to allocate the radio resources reasonably for a variety of transmission services of wireless users. Specificaly, for Non-GBR(non-guranteed bit rate) packet services, it can use an efficient realtime scheduling algorithm to improve the throughput of the system and guarantee the quality of service of every user, and this largely determines the performance of the whole system. In view of the MaxWeight scheduling algorithm provides the optimal throughput for this system, while its delay performance for per-user is very poor, thus this paper use Lyapunov Optimization to design a scheduling algorithm, which makes realtime resource allocation decisions and transmission decisions only based on the current channel states and current queue backlogs, and does not require a prior-knowledge of the channel state probabilities. Theoretical analysis and simulation results show that this scheduling algorithm provides a nice trade-off between throughput and delay, and not only provides optimal throughput, but also significantly reduces the per-user delay.