On ultrasonic communication through metal structure for machine embedded sensing

With the development of machine embedded sensing, wireless communication through metal structures is becoming a critical issue in manufacturing equipment, process monitoring and control. To address it, this paper presents a multi-carrier ultrasonic communication scheme to achieve high-bit data rate for machine embedded sensing through metal structures. To extract the coded information from ultrasonic signal and denoise the signal, a new signal processing method, named multi-scale enveloping symbolic analysis (MuSEnSA), is also formulated by integrating dual-tree complex wavelet packet transform and symbolic analysis. Dual-tree complex wavelet packet transform (DT-CWPT) is firstly investigated by separating multi-carrier signals under noise contamination, given its properties of shift-invariance and flexible time frequency partitioning. A new envelope extraction and threshold setting strategy for selected wavelet coefficients is then introduced to retrieve the coded digital information by taking advantage of symbolic analysis. Numerical and experimental studies are performed to evaluate the effectiveness of the developed signal processing method for injection molding monitoring.