Abstract:
The sixth-generation (6G) mobile communication networks impose unprecedented requirements on spectral efficiency, connectivity scale, and other performance metrics. However, current wireless communication systems predominantly rely on antennas with fixed configurations, whose static radiation characteristics cannot dynamically adapt to varying channel conditions, fundamentally limiting further performance improvements. Reconfigurable pixel antennas, which discretize the radiating aperture into numerous sub-wavelength pixel elements and employ RF switches to enable flexible interconnections among pixels, offer a new degree of freedom for dynamic antenna control and thus present a promising solution for 6G wireless systems.
his talk systematically introduces antenna coding, a novel technique empowered by pixel antennas. We first establish a rigorous electromagnetic-compliant modeling framework based on multiport network theory and beamspace channel representation, which allows us to treat antenna coding, i.e., controlling the switch states, as an independent and controllable degree of freedom, and to jointly optimize it with conventional beamforming and precoding. We then demonstrate the substantial performance gains of antenna coding, supported by either conventional or modern AI based optimization methods, across multiple typical scenarios. Finally, we discuss the potential of antenna coding to integrate with emerging 6G technologies, offering a new electromagnetic reconfigurability paradigm for future wireless networks.