Spatial Chirp-based Joint Optimization for Analog Beamforming in Reconfigurable Intelligent Surface Assisted Terahertz Ultra-wideband Systems

This paper proposes a spatial chirp-based joint optimization method for analog beamforming in RIS-assisted THz ultra-wideband communication systems. The THz band (0.1-10 THz) offers significantly wider bandwidth, enabling Tbps data rates, which are crucial for 6G communications. However, traditional phase shifters are optimized for a single frequency. In ultra-wideband systems, the phase response deviates as frequency varies, leading to beam squinting problem, which significantly degrades system performance. Existing solutions, such as true-time-delay architectures, impose excessive complexity, making them impractical for THz systems. In this work, we propose leveraging polynomial expansion to optimize the spatially dependent phase functions of RIS, Tx, and Rx. We formulate the optimization problem as maximizing system throughput, and develop a gradient ascent-based iterative algorithm to solve it. Simulation results demonstrate that our method outperforms conventional and state-of-the-art approaches in metrics including phase response, throughput, and BER, providing a cost-effective solution for optimizing analog beamforming.