Artificial Intelligence-Based Optimization of Multiband Antennas for Smart Agriculture Applications Using LPWAN Communication

Abstract

This research introduces a compact, AI-optimized multiband antenna specifically designed for Smart Agriculture applications utilizing LPWAN protocols such as LoRa and Sigfox. Focusing on the 868 MHz band—critical for rural and remote IoT deployments—the antenna features a three-layer stacked patch structure to ensure robust multiband performance within a minimized footprint. Artificial Neural Networks (ANNs) and Genetic Algorithms (GAs) are employed to optimize the antenna design parameters intelligently. The ANN is trained on extensive parametric simulations to predict key electromagnetic characteristics, including return loss (S11) and resonant frequencies, while the GA efficiently converges on optimal geometries, significantly reducing simulation overhead. Simulation results demonstrate enhanced performance, with return loss below −19 dB, radiation efficiency exceeding 82%, and tightly controlled bandwidth, ensuring reliable outdoor connectivity. Furthermore, the AI framework incorporates a power estimation model for early-stage VLSI circuits, enabling accurate prediction of energy consumption. This facilitates holistic co-design of antenna and digital subsystems, contributing to extended battery life and scalable sensor node deployment. The proposed methodology supports the development of intelligent, energy-efficient wireless infrastructures for precision agriculture and environmental monitoring.

Keywords: AI-driven multiband antenna design, AI, IOT, VLSI, Smart agriculture connectivity, LPWAN communication, LoRa technology, Neural network optimization, Genetic algorithm-based tuning, Energy-efficient wireless systems.