Passive Internet-of-Things (IoT), a paradigm based on battery-free devices,is a promising technology for realizing large-scale and sustainable wireless connectivity. A key enabler of passive IoT is backscatter communication (BC),which allows devices to communicate without generating their own RF signals. Instead, passive devices convey information by modulating and reflecting incident RF signals. However, the practical deployment of BC systems is fundamentally limited by the severe double path-loss of the backscatter channel and strong direct link interference (DLI) from the carrier emitter(CE) to the reader. These effects lead to extremely weak backscattered signals and impose stringent requirements on receiver hardware, including high-resolution analog-to-digital converters.

This thesis addresses these challenges by leveraging distributed multipleinput multiple-output (MIMO) techniques together with advanced beamforming and detection methods. First, focusing on bistatic BC with two access points, we propose: (i) a novel transmission scheme with a channel estimation protocol; (ii) a transmit beamforming design that suppresses DLI between the CE and the reader; and (iii) a generalized log-likelihood ratio test based detector. The proposed framework significantly reduces the dynamic range requirements at the reader and improves detection performance compared to conventional schemes without interference cancellation.

Extending this setup, the thesis considers BC in distributed MIMO systems, where multiple access points can be flexibly assigned as CEs or readers. In this context, we develop: (i) a joint beamforming and access point role selection framework to illuminate backscatter device and cancel DLI; (ii) a channel estimation method tailored for operation under strong DLI; and (iii) a theoretical analysis including a closed-form expression for the probability of error and a model for quantization noise induced by DLI. The results show that efficient interference management enables reliable operation even with low-resolution analog-to-digital-converters.

Furthermore, we investigate the coexistence of BC with conventional user communications in distributed MIMO systems. Specifically, we propose: (i)a joint optimization framework that maximizes the user performance while satisfying backscatter constraints; and (ii) robust beamforming designs based on the S-procedure to account for imperfect channel knowledge.

In addition, we analyze a practical hardware impairment in distributed MIMO BC systems. In particular, we study: (i) the impact of reciprocity calibration errors on coherent beamforming gain and DLI suppression; and(ii) robust design strategies to mitigate performance degradation under calibration uncertainties.

The theoretical contributions are complemented by experimental validation using a large-scale distributed antenna system, demonstrating substantial gains in interference suppression and signal-to-interference ratio.

Finally, the thesis investigates broad-beam design for multiple antenna systems. In particular, we study the performance of dual-polarized phase-only beamforming for achieving efficient wide-area coverage under both line-ofsight and non-line-of-sight channel conditions. The results provide insights into the use of multiple antenna systems for functionalities such as initial access.

Overall, this thesis provides a comprehensive framework for the design and analysis of BC systems in next-generation wireless networks. By integrating advanced beamforming, detection, signal processing techniques, and AP role selection, it significantly enhances the reliability, scalability, and practical feasibility of passive IoT deployments.