Microwave MIMO E-Nose for Wireless Communication and Selective Detection of VOC Mixtures with Concentration Estimation.

We present the first dual-functional microwave electronic nose (E-nose) that enables wireless communication, VOC mixture detection, and reliable concentration estimation, designed for seamless integration with wireless sensor networks. The proposed E-nose features multiple-input multiple-output (MIMO) antenna system functionalized with molecularly imprinted polymer (MIP) and multiwalled carbon nanotube-based sensing materials for the selective detection of individual or mixed volatile organic compounds (VOCs). We addressed several novel challenges such as managing cross-reactivity under electromagnetic interference with wideband decoupling, employing a dual-branch neural network (NN) with feature prioritization and transducer behavior insights, and optimizing sensor placement for spatial isolation in a compact design. The developed E-nose demonstrated high sensitivity and selectivity for VOC mixtures at room temperature, with detection limits below safety thresholds, ensuring practical applicability. The optimized NN model achieved high predictive accuracy (R2 = 0.982 to 0.991), delivering near-perfect concentration estimations with negligible errors for pure (0.35%) and mixtures of 2-4 VOCs (2.0 to 3.5%). The proposed framework, customizable for detecting diverse VOCs and toxic gases, enables scalable indoor and outdoor air-quality monitoring. Its seamless dual functionality ensures uninterrupted wireless communication services during gas sensing, establishing a new paradigm in advanced sensor technology with microwave MIMO E-Nose systems.