ECAI 2013 - International Conference 5th Edition Electronics, Computers and Artificial Intelligence 27 June -29 June, 2013, Piteşti, ROMÂNIA Evaluation of the beam wondering in free space optics by image analysis Mircea Hulea School of Automatic Control and Computer Science ‘Gheorghe Asachi’ Technical University of Iasi Iasi, Romania mhulea@tuiasi.ro Zabih Ghassemlooy Optical Communications Research Group, Faculty of Engineering and Environment, Northumbria University Newcastle upon Tyne, United Kingdom z.ghassemlooy@northumbria.ac.uk Sujan Rajbhandari University of Oxford, Oxford, United Kingdom, e-mail: sujan.rajbhandari@eng.ox.ac.uk Abstract This paper proposes a simple method to determine the laser beam wandering and scattering in free space optics. The main application of this method is to determine the dimensions of a device required to compensate for beam wandering and scattering due to the air turbulences in FSO communications. The method uses color image analysis for automatic determination of the shape of the laser spot, which provides information about the spot area and the maximum amplitude of beam wandering. Results show that when the spot is photographed on a non-colored paper screen, the method is able to determine the laser spot shape accurately. Moreover, the method is robust to the variation of the screen background illumination. (Abstract) Keywords-FSO link, laser beams scattering and wandering, laser spot shape determination, colour image analysis I. INTRODUCTION Free-Space optical (FSO) communication represents an alternative technology to the radio frequency that uses the unlicensed optical spectrum to offer a high bandwidth for certain applications [1, 2]. The FSO link range is from a few meters to a few kilometers in the outdoor environment. The FSO link is susceptible to a number of environmental effect including fog, smoke and scintillation [3]. Heated air masses in the optical propagation path induce random fluctuation of the air refraction index. This in turn causes a random fluctuation in the optical intensity at the receiver (i.e. scintillation) and beam wandering [4-6]. The scintillation results in a high outage probability in FSO system. There are a number of mathematical models that could be used to predict the scintillation [6-9]. The relation between the link length and the variation of the light intensity variation and of the spot aperture was analyzed experimentally [10, 11]. In [12] the spot displacement in an FSO system is analyzed using image processing. Several methods to improve the FSO link quality and stability such as modulation schemes, channel coding and laser spot trackers are proposed in [13-17]. One of the goals of this study is to determine the minimum size of a tracking device that is able to compensate for beam wandering and scattering. This paper presents a method for determining the area covered by the light spot in order to estimate the size of a tracking device. The unsupervised approach determines the shape of the light spot by analyzing the spot image using a very simple algorithm. The algorithm is relatively simple in successfully detecting the shape of scattered spots due to light propagation through air turbulence. II. EXPERIMENTAL METHOD An FSO link was set-up in a controlled indoor environment. The transmitter consists of a highly collimated narrow divergence (< 5mrad) visible laser. The laser spot after propagation through free space channel of 26 m was photographed and filmed on a non-colored paper screen for different turbulence conditions. Figure 1 presents the optical arrangement used for picturing and filming the laser spot at different distances from the light source. For increasing the light propagation distance in the free space channel we used multiple reflections of the laser beam between two flat mirrors. Using this set-up we obtained images of the displaced and scattered laser spot at distances ranging from 26m to 104m. Images captured were analyzed in order to determine the spot shape as well as the spot displacement. This allowed us to determine the light spot area and the maximum centroid displacement from the position for the case with no turbulence. The method was tested on images of the laser spot taken from different angles and at different distances from the non-colored screen.