Resource Allocation Problems in Wireless Communication Networks

在不久的将来，无线信息网络的预期服务需求超出了这些系统当前的能力。有效地分配有限的频谱将是下一代无线网络设计中的一个关键因素，以满足未来的需求。本文研究了无线通信网络中资源分配问题的两个相关问题。这些是共享本地公共的、空间上可重复使用的无线电信道以及快速变化的拓扑结构的影响。对于第一个问题，我们提出了一种捕获无线网络中的信道分配问题的排队模型。主要的研究目标是对系统性能进行表征，并开发有效共享资源的调度算法。系统的吞吐量和平均分组延迟是两个重要的性能指标。在以前的研究中，我们刻画了上述模型的吞吐量特性，并给出了最大吞吐量调度算法。对于时延，我们得到了串联拓扑网络中的最优调度策略。我们希望通过特别强调可在无线网络的分布式环境中实现的调度策略来扩展上述结果。文中还提出了一个时变拓扑的随机模型。了解时变性对系统稳定性和吞吐量的影响，以及开发有效考虑拓扑信息的传输控制算法是关于改变连通性的主要研究目标。对于单跳网络，已经获得了初步的结果。

The expected service demands in wireless information networks in the near future exceed the current capabilities of these systems. Efficient allocation of the limited spectrum will be a critical factor in the design of the next generation of wireless networks in order to meet the future demands. We study two issues related to the resource allocation problem in wireless communication networks. These are the sharing of the locally common, spatially reusable radio channel and the effect of the rapidly changing topology. Regarding the first issue, we propose a queuing model that captures the channel allocation problem in radio networks. The major research goal is to characterize the system performance and to develop scheduling algorithms for sharing the resource efficiently. The throughput of the system and the average packet delay are the two performance measures of interest. In our previous research we characterized the throughput properties of the above model and we gave maximum throughput scheduling algorithms. Regarding delay, we obtained the optimal scheduling policy in a network with tandem topology. We wish to extend the above results with special emphasis on scheduling policies that are implementable in the distributed environment of a radio network. A stochastic model of time-varying topology is also proposed here. The understanding of the effect of the time-variability on the stability and throughput of the system as well as the development of transmission control algorithms that effectively take into account the topology information are the major research goals regarding changing connectivity. Preliminary results have already been obtained for a single-hop network.