Efficient Recognition for MQAM Signal Using Feature Extraction

In modern communication systems, modulation recognition technology occupies an extremely important position and plays a vital role in both military and civilian fields. However, at present, many technologies have high time complexity and require strictly synchronous down conversion. In this paper, the widely used Multiple Quadrature Amplitude Modulation (MQAM) signal is studied for inter-class classification. Due to the problem that the number of signal elements is different in recognition, based on the high-order cumulant feature and constellation feature theory, this paper proposes a joint recognition scheme combining subtraction clustering. At the same time, for the phase mismatch problem in subtractive clustering, we propose a high-speed and accurate constellation rotation calculation method. Our algorithm can be used for 8QAM, 16QAM, 32QAM, and 64QAM Signal recognition, and the recognition rate is more than 95% under the condition of $\text{SNR}=13\ \text{dB}$ . At the same time, the running time of the proposed algorithm is about 39.1% faster than the current mainstream algorithm, indicating that it can precisely and efficiently identify QAM signals in class.