MIMO Antenna Systems for Integrated Communication and Multi-Analyte VOC Sensing
Abstract
The rapid advancement of wireless communication technologies and the Internet-of- Things (IoT) has opened new avenues for integrating sensing capabilities into existing infrastructure, enabling scalable and cost-effective environmental monitoring solutions. Antenna-based microwave sensors represent a transformative approach, combining sensing and wireless communication functionalities into a single system. This integration not only reduces the spatial and economic footprint of traditional monitoring systems but also leverages widespread wireless networks, such as access points and base stations, to enable real-time, large-scale environmental sensing.
This PhD thesis explores the development of multiple-antenna microwave sensors for environmental monitoring, focusing on gas sensing through simultaneous selective multi-analyte detection and concentration estimation of Volatile Organic Compounds (VOCs) in complex mixtures. Integrating state-of-the-art sensing materials, innovative design strategies, and advanced machine-learning techniques, this research aims to develop highly sensitive, selective, and reliable sensors that operate seamlessly within wireless communication frameworks.
The work begins with the design and optimization of single-antenna sensors using advanced materials, such as Molybdenum Disulfide (MoS2), to achieve high sensitivity for single VOC detection. Building on this foundation, the research incorporates Molecularly Imprinted Polymers (MIPs) and Carbon Nanotubes (CNTs) to enhance the selectivity and sensitivity of individual VOCs. Finally, the research progressed toward developing multi-antenna systems and machine learning-assisted signal processing techniques, enabling the simultaneous detection of multiple VOCs in complex mixtures. These innovations address critical challenges such as cross-reactivity and electromagnetic interference, which are systematically investigated and mitigated through advanced data processing and sensor design.
A key contribution of this research is developing a microwave-based Electronic Nose (E-Nose) system, which integrates advanced sensory components and machine learning models for ultrasensitive detection and precise concentration estimation of multianalyte VOCs, both individually and in mixtures. The system employs a dual-branch Neural Network (NN) model to prioritize features and provide insights into sensor behavior, alongside optimized sensor placement strategies to minimize cross-reactivity and interference in compact designs. To ensure robust performance in real-world environments, low-complexity impedance matching and wideband decoupling techniques are implemented, maintaining uninterrupted wireless communication while enabling reliable sensing.
This research is motivated by the need to address the limitations of traditional gas sensing systems and unlock the potential of IoT-enabled environmental sensing. By bridging the fields of sensing technology, material science, microwave engineering, and machine learning, this research offers a holistic approach to environmental monitoring. The findings presented in this thesis pave the way for scalable, cost-effective, and real-time air quality monitoring systems, enabling large-scale deployment for comprehensive environmental assessment. This has significant implications for public health protection and environmental sustainability. This work underscores the potential of interdisciplinary research to address complex global challenges, leveraging the synergies between emerging technologies to create innovative solutions for a sustainable future.
Keywords: Antenna sensors, microwave sensing, MIMO, gas sensors, volatile organic compounds, VOC, molybdenum disulfide, molecular imprinted polymer, carbon nanotube, electronic nose, machine learning, and neural network.