π€ AI Summary
To address the insufficient robustness of beam tracking in millimeter-wave (FR2/FR3) communications under high-mobility scenarios (e.g., vehicular networks), this paper proposes an adaptive strategy jointly optimizing dynamic spectrum selection and antenna array beam direction. Innovatively, spectrum mobility is modeled as a partially observable Markov decision process (POMDP), and a point-based value iteration (PBVI) algorithm is employed to derive a policy with theoretical performance guarantees. The method maintains high data rates even when user motion deviates significantly from prior assumptions, substantially enhancing beam tracking stability. Numerical experiments demonstrate that the proposed multi-band cooperative beam management scheme achieves significant throughput gains over single-band baseline methods. This work provides a verifiable, deployable, and robust solution for highly dynamic mmWave communications.
π Abstract
The large bandwidths available at millimeter wave (mmWave) FR2 bands (24-71 GHz) and the emerging FR3 bands (7-24 GHz) are essential for supporting high data rates. Highly directional beams utilized to overcome the attenuation in these frequencies necessitate robust and efficient beamforming schemes. Nevertheless, antenna and beam management approaches still face challenges in highly mobile solutions, such as vehicular connectivity, with increasing number of bands. In this work, the concepts of spectrum mobility is studied along with antenna array management in multiple frequencies to improve beamforming under mobility. The spectrum mobility problem aims to select the optimal channel frequency and beam direction in each time slot to maximize data rate. This problem is formulated as a Partially Observable Markov Decision Process (POMDP) and Point-Based Value Iteration (PBVI) algorithm is used to find a policy with performance guarantees. Numerical examples confirm the efficacy of the resulting policy for multiple available frequency bands, even when the user mobility significantly deviates from models assumed during policy generation.