Reconfigurable hybrid beamforming for 6G wireless systems across quasi-far-field and strong near-field regimes

The shift towards sixth-generation (6G) wireless systems calls for a rethink of beamforming designs since at terahertz (THz) frequencies and with large apertures the difference between far-field and near-field propagation is almost gone. In this paper, we propose a methodical comparison framework to evaluate various beamforming methods, including fully digital precoding, fully connected and sub-connected time-delay (TD) and phase-shifter (PS) arrays, delay-phase precoding (DPP), and reconfigurable intelligent surface (RIS)-assisted beamforming. Performance is assessed in terms of spectral efficiency, energy efficiency, and scalability under both propagation regimes. The results show that the performance hierarchy strongly depends on the operating environment. Our findings indicate that the performance scaling order is highly reliant on the operating environment. Fully digital precoding achieves the highest spectral efficiency in all scenarios; however, its relative advantage over the practical architectures reduces from approximately $$16%$$ for near-field narrowband cases to about $$4%$$ for quasi-far-field wideband cases. On the other hand, RIS-assisted beamforming excels in energy efficiency, showing increases of up to $$43.6%$$ over the next-best alternative. Scalability exhibits regime-dependent properties: linear patterns in the quasi-far-field and substantially nonlinear enhancements in the strong near-field. These results may inform the development of efficient and scalable 6G systems.