위상 및 주파수 오프셋에 강인한 개선된 변조 분류 알고리즘

초록

In a non-ideal channel environment where the phase offset and the frequency offset exist, it is important to select features that are robust to performance degradation factors. In this paper, we propose a cumulant based automatic modulation classification algorithm which is independent of the phase and the frequency offset. In order to classify linear digital modulation schemes, e.g., BPSK, QPSK, 8PSK, 16QAM, and 64QAM, we use the cumulants of differential decoded signal as features, and by calculating the euclidean norm of the obtained feature vectors, we perform the modulation classification. Through computer simulation, we compare the performance of the proposed algorithm and the existing algorithm in terms of average classification probability, and show the superiority of the proposed algorithm.

키워드