具有块对角消息拓扑的无线分层网络自由度研究

项目名称： 具有块对角消息拓扑的无线分层网络自由度研究

项目编号： No.61271283

项目类型： 面上项目

立项/批准年度： 2013

项目学科： 无线电电子学、电信技术

项目作者： 刘锋

作者单位： 上海海事大学

项目金额： 70万元

中文摘要： 基于空间自由度分析的无线网络研究已经成为网络信息论的重要内容，但现有文献大多围绕简单的多源多宿单跳信道展开，对于更复杂的网络仍然缺乏了解。我们调研发现，如把单跳网络进行横向扩展将得到多跳级联的分层网络，进行纵向扩展将得到多网并联的单跳分层网络：这两种情况下网络都具有块对角结构的典型消息拓扑。据此，我们拟对该类较为复杂的网络展开研究，采用包括干扰对齐与干扰中和等先进的干扰管理技术来实现空间资源的最大化利用，探索该类网络的自由度问题，阐明先进干扰管理技术的一般机制，揭示建立网络线性可扩展性的一般规律。其重要性在于：基于自由度分析，我们不但能发现网络瓶颈、找到提升系统容量的有效途径，还可以指导信号的设计以及干扰的管理、获得具有特定属性的传输方案。本项目不仅适用于无基础设施的网络，例如船舶通信系统、传感器网络、Ad hoc网络，还可用于有基础设施的无线网络，例如LTE-A系统，具有广泛的应用前景。

中文关键词： 块对角消息拓扑；自由度；干扰对齐；干扰中和；无线分层网络

英文摘要： Wireless network reseach based on spatial degree of freedom (DoF) analysis has been an important content of network information theory. However, most of current works are focused on single hop networks with muti-source and multi-sink. There is still few understanding about more complicated networks. From survey we find that: horizontal expansion of single hop networks will result in cascaded multi-hop layered networks, while vertical expansion of it will obtain a larger single hop network with parallel subnetworks. Both cases share the same message topology structure with block diagonal matrix. Based on this observation, we plan to study such type of complicated networks in this project. Advanced interference management technique such as interference alignment and interference neutralization will be used to maximize the efficiency of spatial resource. The DoF of such networks will be studied, which can help explore the common mechanism of advanced interference management. Finally we want to reveal the general principle of building linear scaling law for these networks. Its importance lies in that: based on DoF analysis, we can not only find the bottleneck of network capacity and break through it with feasible approach, but guide the signal design, interference management and obtain transmission schemes with spec

英文关键词： Block diagonal message topology；degree of freedom；interference alignment；interference neutralization；wireless layered networks