Multicast Outage Probability and Transmission Capacity of Multihop Wireless Networks
Multicast Outage Probability and Transmission Capacity of Multihop Wireless Networks
复制标题
DOI：
10.1109/tit.2011.2146030
复制
发表时间：
2010-02
影响因子：
2.5
通讯作者：
Chun-Hung Liu;J. Andrews
中科院分区：
文献类型：
--
作者：
Chun-Hung Liu;J. Andrews
多播传输（其中相同的数据包必须传送到多个接收器）是传感器和战术网络的一个重要方面，并且与更普遍研究的单播网络相比，具有几个独特的特征。具体来说，这些包括 1) 相同的数据包必须成功传送到多个节点，2) 任何接收器的中断要求数据包至少重新传输到该接收器，以及 3) 多播速率由链路最弱的接收器控制，以便最大限度地减少中断和重传。本文的第一个贡献是开发了一种易于处理的多播模型和吞吐量度量，以捕获多播无线网络中的每个关键特征。我们利用由发射器和接收器的不同泊松点过程（PPP）组成的泊松簇过程（PCP），然后将多播传输容量（MTC）定义为每次传输尝试可实现的最大多播速率乘以解码延迟和多播中断约束下多播簇的最大强度。多播集群是在其上多播数据包的连续区域，并且为了减少中断，可以将其细分为 v 个更小的多播区域。本文的第二个贡献是对该模型的几个关键方面进行了分析，为此我们得出了以下主要结果。假设多播集群中的每个细分区域允许 τ/v 传输尝试，我们表明 MTC 为 θ(ρkxlog(k)vy)，其中 ρ、x 和 y 是 τ 和 v 的函数，具体取决于网络规模和强度，k 是集群中预期接收器的平均数量。我们针对许多感兴趣的机制导出了 {ρ, x, y}，并且还表明适当数量的重传可以显着增强 MTC。
Multicast transmission, wherein the same packet must be delivered to multiple receivers, is an important aspect of sensor and tactical networks and has several distinctive traits as opposed to more commonly studied unicast networks. Specially, these include 1) identical packets must be delivered successfully to several nodes, 2) outage at any receiver requires the packet to be retransmitted at least to that receiver, and 3) the multicast rate is dominated by the receiver with the weakest link in order to minimize outage and retransmission. A first contribution of this paper is the development of a tractable multicast model and throughput metric that captures each of these key traits in a multicast wireless network. We utilize a Poisson cluster process (PCP) consisting of a distinct Poisson point process (PPP) for the transmitters and receivers, and then define the multicast transmission capacity (MTC) as the maximum achievable multicast rate per transmission attempt times the maximum intensity of multicast clusters under decoding delay and multicast outage constraints. A multicast cluster is a contiguous area over which a packet is multicasted, and to reduce outage it can be tessellated into v smaller regions of multicast. The second contribution of the paper is the analysis of several key aspects of this model, for which we develop the following main result. Assuming τ/v transmission attempts are allowed for each tessellated region in a multicast cluster, we show that the MTC is Θ(ρkxlog(k)vy) where ρ, x and y are functions of τ and v depending on the network size and intensity, and k is the average number of the intended receivers in a cluster. We derive {ρ, x, y} for a number of regimes of interest, and also show that an appropriate number of retransmissions can significantly enhance the MTC.