A Study of Error Reduction in Optical Triangulation for Short-Range Intervehicle Positioning

Abstract

The long-existing problem of high positioning inaccuracy by triangulation in the far region in short-range vehicle-to-vehicle (V2V) communication environments is investigated in this study. The triangulation is achieved by sensing angle of arrival with two optical signal-direction sensors developed in our previous works. The measured signal-direction error can be categorized into two different kinds, namely systematic error and random error. The former is mainly incurred by imperfect calibration of hardware and misalignment of signal-direction sensors as well as multipath scattering of the signal. The latter is caused by random noises. Both of them lead to positioning error. By utilizing the specific characteristics pertaining to the triangulation measurements, it is possible to extract the target location and increase the accuracy with an optimization algorithm. With the aid of a general-purpose optimization subroutine together with two proper objective functions developed in this study, this work shows the viability of this idea. With the technique proposed in this paper, precise hardware calibration and careful alignment of signal-direction sensors are unnecessary. The problem of multipath scattering can also be mitigated. This work indicates a possible direction to treat the problem of high inaccuracy in noisy and ill-geometric environments in triangulation measurement. The abovementioned signal-direction sensors are currently constructed for infrared light. By changing the signal sources and detectors, they can also be fabricated for visible light. Such an alteration of transmission medium is beneficial for the currently emerging short-range visible-light intervehicular communications.

Keywords

• Angle of arrival
• geometric dilution of precision (GDOP)
• optical communication
• optimization
• positioning
• signal-direction sensing