Implementation of an Automatic Modulation Recognition System on a Software Defined Radio Platform

This paper discusses the implementation of an automatic modulation classification system on a Software Defined Radio (SDR) platform, devoted to cognitive radio applications. The main objectives are to achieve an optimal realization in terms of the complexity of implementation, and flexibility which is a critical and essential factor in software radio, and to test a digital signal modulation mode identification technique, to distinguish between frequency modulation (2FSK and 4FSK) on the one hand and amplitude and phase modulations (ASK, BPSK, QPSK and 16QAM) on the other hand. The modulation schemes involved in this work are: BPSK, QPSK, QAM, ASK and FSK. The modulation identification algorithm uses the maximum value of the power spectral density of the normalized-centered instantaneous amplitude of the received signal as a metric. The test results for an AWGN channel validated the performance of the proposed architecture.