Sequential Hybrid Beamforming for Radio Stripes

Cell-free (CF) massive multiple-input multiple-output (mMIMO) systems will revolutionize wireless communication by moving away from traditional base stations (BSs). By eliminating conventional cell boundaries, CF mMIMO deploys a large number of distributed access points (APs), enabling seamless connectivity across a wide area, more flexible resource allocation, and enhanced overall system performance. However, CF demands a significant amount of fronthaul signaling overhead due to its inherent nature (star topologies), which limits its practical application. In this regard, radio stripe (RS) is a promising and cost-efficient approach to deal with these issues, improving system efficiency and scalability. Not only RS, millimeter wave (mmWave) communications also represent a significant advancement in the evolution of wireless communication since it achieves multi-Giga bit per second (Gbps) data transfer rates. Still, mmWave requires beamforming to overcome the high path loss (PL) and shadowing effects. In this paper, we propose a new low-complexity hybrid analog-digital beamforming technique that addresses the challenges posed by wideband mmWave CF mMIMO RS-based systems. The APs hybrid equalizer is optimized, using as a metric, the mean squared error (MSE) between the transmitted and the AP received signals. The optimization is done over the radio stripe, i.e., at each AP, only local and side information from the former AP in the stripe is considered. Simulation results have shown that our proposed sequential hybrid algorithm achieves approximately the same performance as the centralized CF counterpart with a significant reduction in the signaling and information exchange requirements.