Modeling mobility and workload for wireless metropolitan area networks
Modeling mobility and workload for wireless metropolitan area networks
复制标题
DOI：
10.1016/j.comcom.2003.10.015
复制
发表时间：
2004-05-20
影响因子：
6
通讯作者：
Steinmetz, R
中科院分区：
文献类型：
--
作者：
Hollick, M;Krop, T;Steinmetz, R
大规模的无线城域网，提供宽带容量，同时支持用户和终端的移动性的研究遭受缺乏现实的移动性和工作负载模型。有一个强烈的需要，这样的模型能够执行健全的模拟支持重要而困难的任务，如网络规划和流量工程。本文提出并研究了一种新的用户移动性真实感建模方法。我们制定了一个分析模型，这是一个混合的经验流动性模型和综合交通模型。该模型明确区分了移动性和交通的影响，以实现更大的灵活性。移动性部分是基于统计分区信息与移动模式的字段数据的组合。这使我们能够预测用户的密度-分为不同的群体-在给定的时间在给定的区域。我们能够在我们的移动模型之上优雅地整合不同的交通特征。用户密度与建模用户组的预测合成流量的组合给出了整个网络的流量和流量波动，从而描述了设想场景的工作负载。我们提出了一个真实的城市的例子，我们的模型的实例化。理论分析和仿真结果表明，该方案具有良好的应用前景.我们的研究结果是，我们的模型能够更精确地覆盖现实世界行为的宏观影响比目前可用的移动性/工作负载模型。(C)2003 Elsevier B. V.保留所有权利。
Research on large-scale wireless metropolitan area networks, which offer broadband capacity while supporting user and terminal mobility suffers from the lack of realistic mobility and workload models. There is a strong need for such models to be able to perform sound simulations supporting important yet difficult tasks like network planning and traffic engineering. In this paper, a novel approach towards realistic modeling of user mobility is proposed and studied. We formulate an analytical model, which is a hybrid of an empirical mobility model and a synthetic traffic model. The model clearly separates the influence of mobility and traffic to allow for greater flexibility. The mobility part is based on the combination of statistical zoning information with field data of movement patterns. This allows us to predict the density of users-classified into different groups-for a given area at a given time. We are able to integrate different traffic characteristics on top of our mobility model elegantly. The combination of user density with the predicted-synthetic-traffic of the modeled user groups gives the traffic and fluctuations of traffic throughout the network, thus describing the workload for the envisioned scenario. We present the instantiation of our model for the example of a real city. Analysis and simulations are provided which show that the proposed scheme is quite prospective. Our findings are, that our model is able to cover the macroscopic effects of real-world behavior more precisely than currently available mobility/workload models. (C) 2003 Elsevier B.V. All rights reserved.