Sampled baseband correlators for in-air ultrasonic rangefinders

摘要

Conventional in-air ultrasonic rangefinders for robotic applications employ inexpensive analog signal processing techniques, based on threshold methods, to perform range measurements. Digital signal processing techniques using correlation methods can provide more accurate measurements, at the expense of greater system complexity and cost. The authors propose using a modified second-order bandwidth sampling technique for making data acquisition concomitant with frequency shift of the sampled narrow-band signals from in-air ultrasonic transducers to their equivalent baseband representations. The reduction of the frequency-analyzing interval allows one to conceive sampled baseband correlators for multiple receiving transducers with a complexity comparable to that of analog receivers. They develop a procedure of sensor calibration that is based on a stochastic model of time-of-flight noise; the effect of the receiver noise on the ranging accuracy is, thus, evaluated separately from the effect, due to environmental changes of the speed of sound. The simulation study and the experimental results obtained with their demonstration system offer promise for multichannel sampled baseband correlators, the time-of-flight estimation accuracy of which is close to the theoretical Cramer-Rao (CR) lower bound for signals in additive white Gaussian noise.