The Research Of Typical Wireless Communication Signal And Biomedical Signal On Technology Of Pattern Recognition

With the rapid development of wireless communication technology, a variety of electronic communications equipment, intelligent electrical equipment, personal portable terminal device, have been rapidly popularized, Many industries are actively involved in applying the latest technology and industry into real life. In the remote medical monitoring system, the information acquired by the physiological parameter acquisition terminals usually need to be transmitted to the remote monitoring center through the communication network. While the electromagnetic spectrum environment becomes more complex, so wireless communication signals in the communication process are often affected by the outside world. So the research on the parameter estimation and pattern recognition of wireless communication signal, automatic diagnosis of biomedical signals have become a hot research topic, it also has important practical significance at the same time.In this paper, we first study the typical communication signal in telemedicine communication network and the common signal modulation recognition method. And the common wireless communication signals such as ZigBee, Wi-Fi, and Bluetooth were simulated and studied, a parameter estimation and recognition method for Bluetooth based on smoothed pseudo Wegener distribution is proposed. The experimental results show that the method is effective.And then, based on the mixed signal pattern recognition under the condition of spectrum coexistence, the theory of mixed signal recognition based on blind source separation is studied. An improved fast independent component analysis algorithm and a hybrid spectral separation algorithm based on constrained independent component analysis are presented, and the validity of the algorithm is proved by experiments. And then extract the high order of the six kinds of signals, four feature parameters are constructed as input feature vectors of Support Vector Machine (SVM)classifier. Finally, the SVM classifier is used for pattern recognition, and the experimental results show that the method is practical and effective.At last, the pattern recognition technology of ECG signal is studied. An automatic diagnosis method based on extreme learning machine is proposed. In order to accurately classify, Wavelet multi-scale decomposition is used in the preprocessing stage, and then the principal component analysis is used to reduce the dimension of data, Feature extraction is based on time domain feature and auto regressive power spectrum feature to construct feature set. Finally, the classification is carried out by extreme learning machine, MIT-BIH arrhythmia database was used to carry out the experiment. Classification experiments on five different types of signals are carried out, the average classification accuracy was 98.18%, the experimental results show that the proposed method is effective and suitable for automatic diagnosis of heart rate disorders.