基于凸优化和遗传算法的分层多播自适应优化

Adaptive layered multicasting based on convex optimization and genetic algorithm

针对无线网络上视频多播效用优化问题,提出了一种新的结合自适应调制编码的分层多播自适应优化方法。具体做法是将视频的各层对应传输于不同的自适应调制编码模式,并对各视频层分配合适的带宽,如果效用函数为凸函数,则系统效用的优化问题转换成凸优化问题,若效用函数无凸性限制,并考虑实际因素,即视频分层编码的层数限制以及有限的、可选择的各层速率之后,新的系统优化问题则通过遗传算法求解。仿真结果证明了该方法的有效性。

Aiming at the utility optimization for wireless video multicast, a new adaptive layered multicast method with AMC（Adaptive Modulation and Coding） was presented. The video layers were transmitted over different AMC modes and allocated appropriate bandwidth. When the utility function was convex, the utility maximization could be converted into a convex optimization problem. When the utility function was not convex constraint, and several practical issues like limited layer numbers and operational layer rates were taken into account, GA（Genetic Algorithm） was proposed to maximize the total utility. The simulation results show that the proposed method is effective.