项目名称: 图信号滤波与估计的理论及应用研究
项目编号: No.61501124
项目类型: 青年科学基金项目
立项/批准年度: 2016
项目学科: 无线电电子学、电信技术
项目作者: 冯辉
作者单位: 复旦大学
项目金额: 21万元
中文摘要: 图信号是一种描述网络信息的新方式。通过图上的变换域分析,图信号处理方法可以有效提取节点信号在不规则拓扑上的关联成分。本项目将以无线传感网为应用背景,研究图信号的滤波和估计问题。提出图滤波器的参数设计和模型复杂度选择方法,探讨边权重对于图信号谱域分布及图滤波的影响。理论成果可用于建立分布式信号估计的图滤波框架,解决无线传感网中拓扑关联信号的分布式估计,并扩展至其他网络信息处理领域。
中文关键词: 图信号;图滤波;分布式估计;传感器网络
英文摘要: Graph signal is a novel description of network information. By analysis on transformation domain of graph signals, the methods of signal processing on graphs can extract the relevance from signals on irregular typologies. We plan to study the filtering and estimation of graph signals from the background of wireless sensor networks (WSNs), including the design of graph filters parameters, model complexity, as well as the impacts of edge weights on spectral distribution and graph filtering. The research results are expected to build the graph-filtering framework of distributed signal processing, which can be applied to solve the distributed estimation on topological related signals in WSNs, and extended to other related fields of network information processing.
英文关键词: Graph Signals;Graph Filtering;Distributed Estimation;Sensor Networks