Algorithm Research And Implementation Of Modulation Recognition Based On Modular RF Instruments

Modulation recognition is process between signal detection and signal demodulation, it can provide information for signal demodulation. In electronic countermeasure(ECM) and civil radio management,modulation recognition has achieved some applications, however, current modulation recognition technology has limitations on recognition performance, recognition range and applications. This article studies modulation recognition algorithms and implement them on hardware platforms. The main contents of this article are as follows:The principles of various modulation signals and the differences in time domin or frequency domin among them are analysed. The algorithm for analog modulation recognition is studied and the process of recognition is given. The principles of dicision tree and support vector machine are studied, AM, FM, DSB, LSB, USB signals are recognized with dicision tree and support vector machine separately. Simulation results show that ratios of correct recognition of five kinds of signals are higher than 95% under the circumstance where signal-to-noise ratio(SNR) is 10dB.The algorithms of digital modulation recognition are studied, including instantaneous feature extraction, clustering algorithm and high-order cumulants. Instantaneous feature extraction method which is based on Hilbert transform is firstly introduced, this algorithm can recognize 2ASK,4ASK, BPSK, QPSK,2FSK and 4FSK signals and the ratios of correct recognition are higher than 90% under the circumstance where SNR is 8dB. Secondly, the principles of clustering algorithm and common clustering algorithms are introduced, and subtractive clustering method is used for modulation recognition. MQAM, MPSK, MASK signals are separately recognized using this method. The principles and nature of high-order cumulants are introduced finally. Many digital modulation signals are recognized by high-order cumulants in circumstances of low signal to noise ratio. The recognition algorithm for OFDM and digital and simulation result is given in this article.Simulation result shows that correct recognition ratios of 4ASK, BPSK, QPSK,2FSK,16QAM,32QAM,OFDM signals are higher than 90% under the circumstance where SNR is 8dB.The implementation of modulation recognition system on NI PXIe modular instruments hardware platforms is studied. The softwares, hardwares and system architecture is introduced in this article. The panels and block digrams of modulation recognition system are designed and modulation recognition is realized. Simulation results show that modulation system can recognize modulated signals correctly and the ratios of correct recognition are higher than 90%.Lastly, the work of this article is summarized and work in the future is prospected.