Spatially Continuous Near-Field Tracking With Sparse Arrays

Abstract

User tracking is a key enabler for delivering situational awareness and applications in next-generation wireless networks. However, challenges arise from the high computational demands and hardware costs of near-field tracking systems. This paper proposes a method for spatially continuous near-field tracking tailored to spatial sparse arrays (SAs). The proposed method enhances tracking performance while effectively reducing computational time by leveraging the extended aperture of the SA and the powerful capabilities of a regression neural network (NN). By effectively mitigating the ambiguity inherent in SA, the proposed method ensures the same level of tracking accuracy as conventional uniform linear arrays with the same aperture size, but at a significantly lower hardware cost. Furthermore, by reducing the required number of antenna elements and incorporating regression NN, the computational burden is significantly alleviated. At the same time, the proposed method further improves tracking accuracy through a spatially continuous solution, achieving both efficiency and precision, emphasizing its practical advantages.