A Comparative Analysis of Key Technologies in Intelligent Signal Processing for 6G
TL;DR
A comparative analysis of key technologies in intelligent signal processing for 6G based on a literature survey indicates that AI-based methods offer superior performance and lower computational latency in complex scenarios, yet they continue to face challenges related to interpretability, generalization, and standardization.
Abstract
— The evolution of sixth-generation (6G) mobile communication systems introduces new challenges for signal processing technologies. The incorporation of emerging techniques such as ultra-massive multiple-input multiple-output (MIMO), reconfigurable intelligent surfaces (RIS), and integrated sensing and communications (ISAC) necessitates a paradigm shift in signal processing from conventional optimization methods toward intelligent approaches. This paper presents a comparative analysis of key technologies in intelligent signal processing for 6G based on a literature survey. Focusing on three core enabling technologies, the study compares traditional optimization-based methods with artificial intelligence (AI)-driven approaches in terms of their performance characteristics and application boundaries in tasks including channel estimation, beamforming, and signal detection. The findings indicate that AI-based methods offer superior performance and lower computational latency in complex scenarios, yet they continue to face challenges related to interpretability, generalization, and standardization. Model-driven deep unfolding, graph neural networks, and generative AI represent important directions for the future development of intelligent signal processing. This paper aims to provide a reference framework for theoretical research and system design in intelligent signal processing for 6G.