Design and Implementation of a Tree-Based Blind Modulation Classification Algorithm for Multiple-Antenna Systems

This paper presents the design and testbed implementation of a tree-based algorithm, which classifies many of the linearly modulated signals, such as binary phase-shift keying (PSK), quadrature PSK (QPSK), offset QPSK, $\pi $ /4-QPSK, 8-PSK, minimum shift keying, and 16-quadrature amplitude modulation. The proposed modulation classification algorithm is applicable to spatially multiplexed multiple antenna systems over frequency-selective fading channels. It works in the presence of timing, phase, and frequency offsets, without having prior knowledge of the channel state information. Classification is performed by utilizing the combined properties of the correlation functions, cyclic cumulant (CC), and cumulant. The correlation function of the received baseband signal exhibits peaks at a particular set of time lag, which represents distinctive features for the modulation formats. The CC uses the position of nonzero cycle frequency of the received baseband signals to classify modulation formats, while the cumulant employs threshold values. To authenticate the proposed algorithm, implementation and measurement are performed in an indoor propagation environment by using a National Instrument testbed setup. The performance of the proposed algorithm is compared with existing methods via Monte Carlo simulations.