| Popis: |
This research proposes and evaluates a feature based modulation classification system designed to discriminate between AM, BFSK, OFDM, DS-CDMA, 4-ASK, 8-ASK, BPSK, QPSK, 8-PSK, 16- PSK, 16-QAM, and 64-QAM signals without a priori knowledge of critical signal parameters, including carrier frequency, symbol rate, or phase offset, among others. The classifier is based on the principles of cyclostationarity and leverages cyclic statistics to make its classification decision. The classification process is performed in a hierarchical process in order to exploit the lower variance of lower order cyclic statistics before making use of the higher order cyclic statistics. The initial classification is based on the estimated Spectral Coherence Function (SOF) of the received signal, followed by estimates of the signal's fourth-through eighth-order Cyclic Cumulants (CC). The performance of the classification system is evaluated under fading flat, two-path, and 20-path transmission channels, and multiantenna combining methods are exploited to increase the system performance. |
