Study in Analysis Method of Nonstationary Signals

在自然现象和社会现象中，大多数观测到的信号都是非平稳的。然而，传统的信号分析方法是建立在平稳条件的假设基础上的，对于实际信号，这些方法所带来的结果的可靠性并不令人满意。另一方面，由于生物系统对非平稳信号的检测、提取和分析具有灵活性和精确性，因此对信号分析技术的能力要求很高。研究人员研究了一种基于信号时变所包含的信息来分析非平稳现象的新方法，并提出了一种新的非平稳过程分析方法，该方法应用于包括人类睡眠脑电图信号在内的生物系统实际信号的分析。将调制作为一种非平稳信号，对扩频调制方案及其在个人移动通信业务中的应用进行了基础研究。

Most of observed signals in natural and social phenomena are nonstationary. However the conventional signal analysis methods are based on an assumption of stationary condition, so that the reliability of results brought by these methods are not satisfied for practical signals. On the other hand, high capability of signal analysis technique is required as bio-systems have flexibility and preciseness for detection, extraction, and analysis corresponding to nonstationary signals.The researchers studied a novel approach to nonstationary phenomena based on the information contained in time-variations of signals, and proposed a new method for nonstationary processes supported by this Grant, The new analysis method applied to practical signals of bio-system including Electroencephalogram signal in human sleep. Considering modulation as one of nonstationary signals, the researchers studied also basic investigation of Spread Spectral (SS) modulation scheme and application to personal mobile communication services.