项目名称: 基于矩阵秩极小化模型的无线传感网络定位算法研究
项目编号: No.11301022
项目类型: 青年科学基金项目
立项/批准年度: 2014
项目学科: 数理科学和化学
项目作者: 罗自炎
作者单位: 北京交通大学
项目金额: 22万元
中文摘要: 无线传感网络定位是无线传感网络技术的一个极具挑战性的核心基础问题。它是信息科学、无线通信、计算机科学、运筹学、环境科学、以及工程领域等共同关心的一个热点研究课题,并在国防军事、国家安全、交通管理、灾害预测、医疗卫生、城市信息化建设等多个领域有着广泛应用。本项目旨在将流行的矩阵优化理论与算法,结合网络定位问题独特的结构与性质,研究基于矩阵秩极小化模型的无线传感网络定位算法。其研究内容包括:从近几年蓬勃发展的低秩矩阵恢复的角度,分析网络定位相应的两类特殊矩阵秩极小化模型及其松弛模型的约束系统和解集的理论性质并建立精确松弛理论;为网络定位设计高效、稳定、快速收敛的优化算法;将新算法应用于实际定位问题。本项目的实施不仅能够丰富低秩矩阵恢复理论,而且为无线传感网络定位提供新理论与新算法,同时也可促进信息科学、无线通信、计算机科学与最优化的交叉与融合,具有重要的科学意义和实用价值。
中文关键词: 矩阵秩极小化;半定规划;无线传感网络定位;松弛策略;优化算法
英文摘要: Wireless sensor network localization (WSNL for short) is one of challenging, fundamental and key problems in wireless sensor networking. It is a popular research topic of common interest in information science, wireless communication, computer science, operations research, environment science and engineering. And it has been widely applied to areas of National Defence Military Affairs, National Safety, Traffic Control, Disaster Prediction, Health Care, and City Informatization Construction, etc. This project aims to study the matrix rank minimization based methods for WSNL by employing the matrix optimization theory and methods, together with the special structure of WSNL. The research contents are as follows: analyze the theoretical properties of constraint system and solutions to two specific matrix rank minimization models and their relaxations related to WSNL from the low-rank matrix recovery perspective; design highly efficient, stable and fast convergent optimization algorithms for WSNL; use the proposed algorithms to solve practical WSNL problems. This project has scientific significance and great practical values for the refinement of general low-rank matrix recovery theory, for supplying WSNL with new theory and methods, and also for the promotion of the crossing of information science, wireless commun
英文关键词: Matrix rank minimization;Semidefinite programming;Wireless sensor network localization;Relaxation strategies;Optimization algorithms