无线认知传感器网络中时空相关频谱感知算法研究

项目名称： 无线认知传感器网络中时空相关频谱感知算法研究

项目编号： No.61501224

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： 王天荆

作者单位： 南京工业大学

项目金额： 19万元

中文摘要： 频谱感知是实现认知无线电技术的前提和基础，它对缓解频谱资源短缺问题有重要意义。本项目研究无线认知传感器网络中时空相关频谱感知算法，以解决无线传感器网络频谱资源短缺问题，增强网络数据传输量。本项目以频谱感知信号的时空相关性为主线，研究窄带和宽带两种场景下低能耗的频谱感知算法，实现动态频谱接入和频谱资源共享，提高频谱利用率。针对窄带频谱，根据基于独立成份分析的主用户定位算法，提出序贯协作频谱感知算法，以能量有效的方式实现快速、高性能的频谱检测；针对宽带频谱，单节点利用压缩感知对非理想的块稀疏信号进行盲稀疏度下序贯采样和聚类重构，并据此建立多节点频谱感知信号集合的时空相关性模型，提出基于分布式压缩感知的协作频谱感知算法，实现高检测率、低复杂度和低能耗的协作频谱检测。

中文关键词： 无线认知传感器网络；频谱感知；时空相关性；压缩感知；分布式压缩感知

英文摘要： Spectrum sensing is the precondition and basis for implementation of cognitive radio technology, which is very important to alleviate the problem of spectrum scarcity. In order to solve the issue of spectrum scarcity and strengthen the data transmission in wireless sensor networks, new spectrum sensing algorithms in wireless cognitive sensor networks are proposed. Regarding the temporal-spatial correlation of sensing signals as a mainline, the spectrum sensing algorithms with low energy cost for narrow band and wide band spectrum are discussed to realize dynamic spectrum access and spectrum coexistence and improve the spectral efficiency. According to the location algorithm of primary user based on independent component analysis, a sequential cooperative spectrum sensing algorithm for narrow band spectrum is designed to realize quick and high performance spectrum sensing by the way of energy efficiency. Under blind sparsity, each sensor node sequentially samples the nonideal block sparsity signals of wide band spectrum and reconstructs those signals with clustering algorithm. A temporal-spatial correlation model of sensing signal assemble is build to propose collaborative spectrum sensing algorithm based on distributed compressed sensing, which realizes collaborative spectrum detection with high detection probability, low complexity and low energy cost.

英文关键词： wireless cognitive sensor networks; spectrum sensing;temporal-spatial correlation;compressed sensing;distributed compressed sensing