Abstract

The task of Application Layer Multicast (ALM) routing is to establish a proper routing tree on application layer network, which transmits data packages from a source to multiple destinations. ALM routing tree is instable due to the end-hosts’ departure or failure. In view of this, we formulated the ALM routing as a multi-objective optimization problem, where the aim is to optimize the ALM routing tree’s topological structure for minimizing its transmission delay and instability simultaneously. A novel encoding-free non-dominated sorting genetic algorithm is proposed for solving our formulated optimization problem. For achieving encoding-free, genotypes are directly represented as tree-like phenotypes in our proposed algorithm. Accordingly, the genetic operators acting on genotypes, like crossover and mutation, need to be redesigned to adapt the tree-like genotypes. The worst-case time complexity of the proposed algorithm is theoretically analyzed and exhaustive simulation results manifest that our proposed algorithm is capable of obtaining high-quality Pareto front. Besides, several interesting issues, such as how to select the final solution out of the obtained Pareto front and the reason why GP is not used, are also discussed.

中文翻译：

---

通过免编码非支配排序遗传算法优化应用层组播路由树

摘要

应用程序层多播（ALM）路由的任务是在应用程序层网络上建立适当的路由树，该树将数据包从源传输到多个目标。由于最终主机的离开或故障，ALM路由树不稳定。有鉴于此，我们将ALM路由公式化为一个多目标优化问题，目的是优化ALM路由树的拓扑结构，以同时最小化其传输延迟和不稳定性。为解决我们提出的优化问题，提出了一种新颖的无编码非支配排序遗传算法。为了实现无编码，在我们提出的算法中，基因型直接表示为树状表型。因此，作用于基因型（例如交叉和突变）的遗传算子 需要重新设计以适应树状基因型。从理论上分析了该算法在最坏情况下的时间复杂度，详尽的仿真结果表明，该算法能够获得高质量的帕累托前沿。此外，还讨论了一些有趣的问题，例如如何从获得的Pareto前沿中选择最终解决方案以及不使用GP的原因。