The Cao method for determining the minimum embedding dimension of sea clutter

To verify whether sea clutter has chaotic character or not, the minimum embedding dimension and time delay in phase space reconstruction are needed. This paper discusses two defects of False Nearest Neighbors (FNN) for calculating embedding dimension: noise sensitivity and effect of subjective parameter. It was confirmed that Cao method could overcome these limitations and settle on a suitable embedding dimension of time series and distinguish deterministic signals and stochastic signals clearly. It is concluded that sea clutter is composed of stochastic component and determined component together.