LoRD-Net: Unfolded Deep Detection Network With Low-Resolution Receivers

The need to recover high-dimensional signals from their noisy low-resolution quantized measurements is widely encountered in communications and sensing. In this paper, we focus on the extreme case of one-bit quantizers, propose a deep detector entitled LoRD-Net for recovering information symbols measurements. Our method model-aware data-driven architecture based unfolding first-order optimization iterations. has task-based dedicated underlying signal interest without requiring prior knowledge channel matrix through which are obtained. proposed much fewer parameters compared black-box networks due incorporation domain-knowledge design its architecture, allowing it operate fashion while benefiting flexibility, versatility, reliability model-based methods. operates blind fashion, requires addressing both non-linear nature data-acquisition system as well identifying proper objective recovery. Accordingly, two-stage training LoRD-Net, first stage form process unfold, latter trains resulting model an end-to-end manner. We numerically evaluate receiver recovery wireless demonstrate that hybrid methodology outperforms state-of-the-art methods, utilizing small datasets, order merely ? 500 samples, training.