一种无线传感网中的盲信号分离新方法

无线传感器网有资源受限的特点，所以要在无线传感器网络中在线地实现复杂的盲信号分离给传感器网络的计算资源和网络寿命提出了巨大挑战。近年来已经有许多学者在研究基于Renyi熵准则实现盲信号分离，并且取得不少进展。尽管Renyi熵作为Shannon熵的一般形式表现出更加好的统计适应性，但同时Renyi熵的多峰性和复杂性对寻优算法提出高的要求，结合无线传感器网络特点，引入遗传算法和正交滤波方法来搜寻Renyi熵势能最小点从而实现盲信号分离，不仅大大减少算法复杂度，而且克服了基于牛顿或自然梯度类寻优算法在传统独立成分分析中容易陷入局部最优的弱点并且解决了“顺序模糊”问题。

A new Blind Source Algorithm for Wireless Sensor Network

Wireless sensor network is characterized by resources limited problems,so to implement an on-line Blind Source Algorithm in WSN challenges computing resources and network lifetime.In recent years,many scholars focus on Renyi's entropy as a new novel criterion solving the BBS problems.Although Renyi's entropy which bears much complicated statistic characteristics determines a better performance on measuring independence of signals,meanwhile its multimodality and complexity require a more powerful searching o...