Network Coding Based Link Scheduling for QoS Provisioning in Multi-radio and Multi-channel Wireless Mesh Networks
Network Coding Based Link Scheduling for QoS Provisioning in Multi-radio and Multi-channel Wireless Mesh Networks
复制标题
DOI：
10.4304/jcm.7.12.899-908
复制
发表时间：
2012-01
期刊：
影响因子：
--
通讯作者：
Ya-Bo Hu;Fangmin Li;Xinhua Liu
中科院分区：
文献类型：
--
作者：
Ya-Bo Hu;Fangmin Li;Xinhua Liu
本文通过应用一种新颖的网络编码技术，解决了具有多无线电多通道的无线网状网络 (WMN) 的服务质量 (QoS) 改进问题。众所周知，有线网络上的网络编码能够以传统路由无法达到的速率实现连接。然而，无线网络的特性（例如，全向传输、相消性干扰、每个节点单个收发器、有限能量）修改了网络编码的公式，并偏离了有线网络中使用的经典网络编码方法。在本文中，我们研究了多无线电和多通道WMN 上考虑QoS 要求（即数据包传送率、数据包延迟）的全网数据传输问题。为了解决这个问题，我们首先引入一种网络编码方法，即 COPE，用于提高单播无线网络的网络范围吞吐量，然后提出给定多无线电和多信道 WMN 的整数线性规划公式，用于解决网络编码流量、路由、QoS 要求和调度优化。所提出的分析公式通过结合网络编码策略和无干扰调度，提高了网络范围的吞吐量，同时满足广义的 QoS 要求。我们在 16 节点图拓扑和 32 节点随机拓扑网络中的评估表明，与编码无关的路由策略相比，了解网络编码的路由选择策略会带来更高的端到端吞吐量。
This paper addresses Quality-of-Service (QoS) improvement in wireless mesh networks (WMNs) with multi-radio multi-channel by applying a novel network coding technique. It is known that network coding over wired networks enables connections at rates that cannot be achieved by traditional routings. However, the properties of wireless networks (e.g., omnidirectional transmissions, destructive interference, single transceiver per node, finite energy) modify the formulation of network coding and deviate from the classical network coding approach used in wired networks. In this paper, we investigate the problem of network-wide data transmission under the consideration of QoS requirements (namely packet delivery ratio, packet delay) over multi-radio and multi-channel WMNs. To solve this problem, we first introduce a network coding method, namely COPE, which is used to increase network-wide throughput for unicast wireless networks, and then present an Integer Linear Programming formulation of a given multi-radio and multi-channel WMN for addressing network coding traffic, routing, QoS requirements and scheduling optimizations. The proposed analytical formulation increases the network-wide throughput while satisfying the generalized QoS requirements by combining network coding strategy and interference free schedules. Our evaluations both in 16-node graph topology and 32-node random topology network show that a route selection strategy that is aware of network coding leads to higher in end-to-end throughput when compared to coding-oblivious routing strategies.