RRAM-Based Nonlinear Precoding with Linear Complexity and Quantization Analysis

In multi-user communication systems, nonlinear precoding generally exhibits higher throughput than linear pre-coding, however at the cost of higher computation complexity. With the increasing number of antennas and users, the realization of nonlinear precoder at the base station is even more challenging. To address this issue, we propose a new system architecture that employs resistive random-access memory (RRAM) circuits to reduce the computation complexity of the nonlinear Tomlinson-Harashima precoding (THP) to a linear scale. We present a computation-constraints principle for designing RRAM circuits to perform nonlinear operations and construct an LQ decomposition RRAM circuit. Since the conductance of memristor is quantized, we perform the bit precision analysis and derive the lower bound of the Signal to Interference plus Noise Ratio (SINR). Our analysis indicates that at a high Signal to Noise Ratio (SNR) or with a large number of antennas, each 1 bit increase in bit precision brings a 6 dB improvement in SINR. Simulation demonstrates the feasibility and accuracy of the RRAM-based circuit and our theoretical results. Our work proves that the RRAM array holds significant potential for implementing high-complexity nonlinear precoding algorithms and may offer a promising solution to meet the demands of future communication.