Future communication system requires large bandwidth to achieve high datarate up to multigigabit/ sec, which makes analog-to-digital (ADC) become a keybottleneck for the implementation of digital receivers due to its highcomplexity and large power consumption. Therefore, monobit receivers for BPSKhave been proposed to address this problem. In this work, QPSK modulation isconsidered for higher data rate. First, the optimal receiver based on monobitADC with Nyquist sampling is derived, and its corresponding performance in theform of deflection ratio is calculated. Then a suboptimal but more practicalmonobit receiver is obtained, along with iterative demodulation and smallsample removal. The effect of the imbalances between the In-phase (I) andQuadrature-phase (Q) branches, including the amplitude and phase imbalances, iscarefully investigated too. To combat the performance loss caused by IQimbalances, monobit receivers based on double training sequences are proposed.Numerical simulations show that the low-complexity suboptimal receiver suffersonly 3dB signal to noise ratio (SNR) loss in AWGN channels and 1dB SNR loss inmultipath static channels compared with the matched filter based monobitreceiver with full channel state information (CSI). The impact of the phasedifference between the transmitter and receiver is presented. It is observedthat the performance degradation caused by the amplitude imbalance isnegligible. Receivers based on double training sequences can efficientlycompensate the performance loss in AWGN channel. Thanks to the diversityoffered by the multipath, the effect of imbalances on monobit receivers infading channels is slight. I
展开▼
