With the continuous developing of communication and the complexity its environment, the traditional technology of reconnaissance and interference is hard to meet the necessity of communication counterwork. This paper introduces a concept of blind reconnaissance and blind interference, of which the kernel is the compression of the signal data to realize its feature vector for its analysis, matching, reconstruction and interference without any prior knowledge.The paper brings forward a kind of implementation for blind reconnaissance and blind interference, also researched and simulated are the methods feature vector extraction and character pattern recognition. In the implementation, based on the current theory of local wave decomposition, the generation of envelope curves is improved by which the extraction of instantaneous signaling parameter is accurately achieved. Then the signal feature extraction is implemented by Singular value decomposition (SVD), and the feature matching is achieved by support vector machine (SVM). The simulation results obtained shows the high identification of this method based on a few samples without being influenced by signal modulating. |