Parasitic MIMO Beamforming for Multi-Active Multi-Parasitic Antenna Arrays with Binary Control

In 6G, MIMO dimensions continue to scale, yet the increased cost, power consumption, and hardware complexity associated with growing RF chains limit practical deployment. Parasitic antennas offer a promising alternative that can add spatial degrees of freedom and array gain without a proportional increase in RF chains. From a communication perspective, prior work on parasitic antennas has primarily focused on adjusting continuous reactance values using varactors, but such varactor-based tuning has increased cost and complexity in the analog control and practical RF circuit design. This paper proposes a multi-active multi-parasitic antenna (MAMP) architecture with binary controllers, where each parasitic element operates in one of two discrete reactance states. To validate the practicality of the system, we experimentally identify array geometries that best match the actual radiation patterns with those of the mathematical model through HFSS simulations. We express the induced current vector as a quadratic function of the binary state vector, and propose a pair of discrete reactance values that minimize the relative error of the proposed model while being implementable with off-the-shelf RF components. With these results, we develop two transmit beamforming codebook designs based on the generalized Lloyd algorithm. The first design exhaustively searches for all possible binary combinations to find the optimal solution, representing the theoretical upper limits of our framework. The second design leverages eigenvalue perturbation to significantly reduce computational complexity, making it suitable for online adaptation. Extensive simulations under various channel scenarios demonstrate that the proposed codebook designs enable MAMP with only few active antennas to achieve beamforming performance comparable to fully active antenna arrays with significantly more active antennas.

Source: arXiv:2607.08624v1 - http://arxiv.org/abs/2607.08624v1 PDF: https://arxiv.org/pdf/2607.08624v1 Original Link: http://arxiv.org/abs/2607.08624v1