调度的目标是尽可能降低语音分组时延,同时最大化数据业务的总吞吐量,采用串行调度的方法:先进行语音调度,再进行数据调度。通过一个可调参数Vmax体现语音和数据业务在资源分配上的折衷。算法设计时考虑了可变分组长度给调度带来的影响与约束,提出了一种载波补偿方案以保证在每个调度帧内传输尽可能多的分组,并利用了语音业务特性。仿真结果表明:与Ahmed和Gkkw算法相比,该算法降低了语音分组时延。 The goal of scheduling is to reduce voice packet delay as much as possible while maximizing the total throughput of data services. A serial scheduling method is adopted: voice scheduling is first performed, and then data scheduling is performed. Vmax, an adjustable parameter, reflects the trade-off of voice and data services in resource allocation. The algorithm design considers the influence and constraint of variable packet length on the scheduling. A carrier compensation scheme is proposed to ensure that as many packets can be transmitted in each scheduling frame and the voice service features are utilized. Simulation results show that compared with Ahmed and Gkkw algorithms, this algorithm reduces the voice packet delay.