Performance Analysis of an Acoustic Relative Positioning System

In this work, the performance of an acoustic relative positioning system is studied, by a Monte-Carlo analysis, with the aim of determining the obtained accuracy and precision. The system has been designed in order to be used in the relative positioning estimation of mobile objects which are not restricted to a particular environment and cannot depend on any external infrastructure. In the proposed system, the spatial relations among objects are computed by only using acoustic emissions with the aim of measuring the round-trip-time-of-flight of the acoustic signals emitted by every object. Then, using the data collected in the system, object positions are computed using the multidimensional scaling technique, and a closer solution is achieved by least-square algorithms. In this performance analysis, a Gaussian distribution of the errors has been considered. The obtained results show that a sub-centimeter precision and accuracy can be obtained considering a centimeter error magnitude in the distance measurements with the proposed ranging mechanism.