Dual-mode blind decision feedback equalizer for high-order QAM multi-domain measurement

Based on real-time multi-domain signal analysis architecture for high-order quadrature amplitude modulation (QAM) communication, dual-mode blind adaptive decision feedback equalization (DFE) scheme for high-order QAM real-time multi-domain analysis and specifications measurement is presented. The proposed two-phase scheme consists of the generalized multi-modulus algorithm (GMMA) for initial convergence during blind startup, and the decision-directed least mean square algorithm (DDLMS) which is used to enhance the performance. The two operation modes are automatic switched according to the equalizer output error level. Combined with step-size parameter variation online in each mode, the blind equalization solution can converge rapidly with low residual tap fluctuation noise. The simulation results show that the proposed techniques converge quickly and reduce the residue mean squared error (MSE) caused by the propagation channel to -40 dB. On the other hand, the proposed algorithm is practically implemented in the communication testing system. The measurement error of EVM is less than 2% for 6.4 MSps 256QAM in the 20 MHz real-time analysis bandwidth. For these considerable merits, the scheme is quite suitable for high-resolution communication measurement instruments design.