Fair Allocation of Subcarrier and Power in an OFDMA Wireless Mesh Network
Fair Allocation of Subcarrier and Power in an OFDMA Wireless Mesh Network
复制标题
DOI：
10.1109/jsac.2006.881628
复制
发表时间：
2006-11
影响因子：
16.4
通讯作者：
Ki-Dong Lee;Victor C. M. Leung
中科院分区：
文献类型：
--
作者：
Ki-Dong Lee;Victor C. M. Leung
针对正交频分多址无线Mesh网络，提出了一种新的公平调度方案，该方案公平地为Mesh路由器和Mesh客户端分配子载波和功率，以最大化纳什讨价还价解的公平性准则.在WMN中，由于并非调度所需的所有信息都在中央调度器处可用（例如，MR），基于在每个节点处本地可用的有限信息，在分布式调度中涉及MR和尽可能多的网格客户端是有利的。而不是解决一个单一的全球控制问题，我们分层解耦的子载波和功率分配问题成两个子问题，其中MR分配组的子载波的网状客户端，每个网状客户端分配其子载波之间的发送功率到每个传出链路。分别用非线性整数规划和非线性混合整数规划来描述这两个子问题。提出了一种简单有效的求解算法。此外，通过将网格客户端的问题转化为一个时分调度问题，得到了一个封闭形式的解决方案。大量的仿真结果表明，该方案提供了公平的机会，各个用户（网格客户端）和可比的整体端到端速率时，网格客户端的数量增加
This paper presents a new fair scheduling scheme for orthogonal frequency-division multiple-access-based wireless mesh networks (WMNs), which fairly allocates subcarriers and power to mesh routers (MRs) and mesh clients to maximize the Nash bargaining solution fairness criterion. In WMNs, since not all the information necessary for scheduling is available at a central scheduler (e.g., MR), it is advantageous to involve the MR and as many mesh clients as possible in distributed scheduling based on the limited information that is available locally at each node. Instead of solving a single global control problem, we hierarchically decouple the subcarrier and power allocation problem into two subproblems, where the MR allocates groups of subcarriers to the mesh clients, and each mesh client allocates transmit power among its subcarriers to each of its outgoing links. We formulate the two subproblems by nonlinear integer programming and nonlinear mixed integer programming, respectively. A simple and efficient solution algorithm is developed for the MR's problem. Also, a closed-form solution is obtained by transforming the mesh client's problem into a time-division scheduling problem. Extensive simulation results demonstrate that the proposed scheme provides fair opportunities to the respective users (mesh clients) and a comparable overall end-to-end rate when the number of mesh clients increases