Deteksi Automatis Skema Modulasi Sinyal OFDM menggunakan Ciri Statistik dan Klasifikasi PSO

Modulation recognition of the detected signal is one of important topics on intelligent receiver system used for military and commercial applications. (Therefore) This research explored the classifications of the OFDM signal modulation scheme namely QPSK, 16-QAM and 64-QAM. The OFDM signal was simulated to pass through frequency selective fading channel and additive white gaussian noise. The classification system was developed using statistical feature extraction with a decision diagrams (tree diagram) as a classifier optimized by PSO algorithm. The increasing number of OFDM symbols in the classification process that applied a voting system improved the accuracy of the classification of each modulation scheme. The optimum accuracy of the classification had been obtained when five OFDM symbols were applied in the classification scenario. The accuracy was 100% for QPSK classification, 99 % for 16-QAM classification and 96 % for 64-QAM classification on the minimum SNR accepted by the receiver of a system that applied a standard WiMax IEEE 802.16e.