Wireless Gesture Recognition by Commercial LTE Signal with Inter-Cell Interference Mitigation

Device-free wireless sensing (DFWS) through opportunistic signals has attracted growing interest, primarily due to its unparalleled convenience. Among the various signals available, the 4G long-term evolution (LTE) signal stands out as an excellent candidate for wireless sensing because of its broad coverage and always-on cell-specific reference signal (CRS). However, a unique challenge arises when employing commercial cellular signals, the presence of inter-cell interference (ICI) will seriously degrade wireless sensing performance. In this paper, we introduce a gesture recognition scheme that leverages commercial LTE signals and addresses ICI mitigation. Firstly, we analyze the origins and impact of ICI according to the physical layer resource allocation mechanisms in LTE. Following this, we present a spatial channel estimation algorithm for both the serving cell and neighbor cells under ICI, and then suppress the ICI through null space projection. To validate the effectiveness of our proposed scheme, we developed a prototype LTE receiver and conducted real-time gesture recognition tests. The experimental results demonstrate that the proposed method can achieve high accuracy and outperform other interference handling algorithms.