An integrated deep aliasing spectrum signal parameter identification method

Abstract

This invention discloses an integrated method for identifying parameters of deeply aliased spectral signals. This method designs a three-layer deep decomposition structure, simultaneously estimating the category, location, and power of aliased spectral signals. Traditional NMF (Non-Multi-Functional Mechanism) is a single-layer decomposition, and the decomposition results are not unique. In the overall objective function construction stage, this invention employs a deep NMF approach, decomposing signal parameters layer by layer to identify signal type, power estimation, and location. Different layer factors are designed to give each layer different feature representations, allowing each layer to obtain corresponding signal feature parameters. Using the technical solution of this invention can effectively improve the accuracy and robustness of signal identification, optimize spectrum resource management, and enhance the overall performance of wireless communication systems.

Images