认知WSN中基于能量有效性自适应观测的梯度投影稀疏重构方法

发布时间：2018-06-25 09:01

本文选题：认知无线传感器网络 + 能量有效性　；参考：《电子与信息学报》2014年01期

【摘要】：针对认知无线传感器网络中传感器节点侧的模拟信息转换器对本地感知数据进行稀疏表示与压缩测量,该文提出一种基于能量有效性观测的梯度投影稀疏重构(GPSR)方法。该方法根据事件区域内认知节点对实际感知到的非平稳信号空时相关性结构,映射到小波正交基级联字典进行稀疏变换,通过加权能量子集函数进行自适应观测,以能量有效的方式获取合适的观测值,同时对所选观测向量进行正交化构造测量矩阵。汇聚节点采用GPSR算法进行自适应压缩重构。仿真比较了GPSR自适应重构与正交匹配追踪(OMP)重构算法。仿真结果表明,在压缩比小于0.2的区域内,基于能量有效性观测的GPSR自适应重构效果优于传统随机高斯测量信号重构。在相同节点数情况下,GPSR自适应压缩重构方法在低信噪比区域内具有较小的重构均方误差,且该方法所需观测数明显低于随机高斯观测,同时有效保障了感知节点的能耗均衡。
[Abstract]:In this paper, a gradient projection sparse reconstruction (GPSR) method based on energy efficiency observation is proposed for the sparse representation and compression measurement of local perceptual data by analog information converters at the node side of the sensor in cognitive wireless sensor networks. According to the space-time correlation structure of the perceived non-stationary signals in the event region, the proposed method is mapped to the wavelet orthogonal basis cascade dictionary for sparse transformation, and the weighted energy subset function is used for adaptive observation. At the same time, the measurement matrix is constructed by orthogonalization of the selected observation vector. The convergent node adopts GPSR algorithm for adaptive compression and reconstruction. The GPSR adaptive reconstruction algorithm and the orthogonal matching tracking (OMP) reconstruction algorithm are simulated and compared. The simulation results show that in the region with compression ratio less than 0.2, the adaptive reconstruction based on energy efficiency observation is superior to that of conventional random Gao Si signal reconstruction. In the case of the same number of nodes, the reconstruction mean square error is smaller in the low SNR region, and the number of observations required by the method is obviously lower than that of the random Gao Si observation, and the energy balance of the perceptual node is effectively guaranteed.
【作者单位】：杭州电子科技大学通信工程学院;
【基金】：国家自然科学基金(61102066,61100044,61001133) 浙江省自然科学基金(LY12F01007)资助课题
【分类号】：TP212.9;TN929.5

【参考文献】