Mitigating Quantization Error in mmWave Phased Antenna Arrays

Sammanfattning

Beamforming is a key technology in the 5G mmwave communication system, which can ensure the reliability and efficiency of communication by precisely controlling the signal direction of phased array antennas. Among the digital, analog, and hybrid beamforming techniques available, hybrid beamforming stands out as a balanced solution, offering optimal performance while maintaining cost-effectiveness for large-scale production. Digital attenuators and phase shifters are typically employed for collimated beam formation, but the quantization inherent to these components introduces errors that result in parasitic sidelobes, pointing inaccuracies, and gain loss. Various methods have been proposed to alleviate quantization errors, encompassing rounding-off to various dithering methods. This study compares different dithering methods against basic rounding techniques to identify the optimal dithering strategy for minimizing pointing errors. Dithering methods are applied solely to phase quantization, as amplitude quantization errors are too intricate to resolve through randomization alone. Unlike most studies that address single quantization errors, this research evaluates the cumulative effects of both amplitude and phase errors. Phase errors are shown to impact beam pointing and sidelobe levels significantly, whereas amplitude errors primarily influence sidelobe intensity. The initial phase of the study establishes optimized parameters for phase quantization using leading dithering methods. Subsequently, we employ advanced optimization algorithms—including Quantum Particle Swarm Optimization, Binary Particle Swarm Optimization, and the newly proposed Zebra Optimization Algorithm—to refine the amplitude quantization. This dual approach synergistically reduces the overall quantization impact on sidelobe levels, enhancing beamforming performance. The efficacy of these methods is validated through simulations with large linear phased array antennas and experimental trials using smaller arrays on an SDR-based mmWave test platform, aiming to minimize both sidelobe levels and pointing inaccuracies.