We present the performance benefits of differential phase-shift keying (DPSK) modulation in eliminating influence from atmospheric turbulence, especially for coherent free space optical (FSO) communication with a high communication rate. Analytic expression of detected signal is derived, based on which, homodyne detection efficiency is calculated to indicate the performance of wavefront compensation. Considered laser pulses always suffer from atmospheric scattering effect by clouds, intersymbol interference (ISI) in high-speed FSO communication link is analyzed. Correspondingly, the channel equalization method of a binormalized modified constant modulus algorithm based on set-membership filtering (SM-BNMCMA) is proposed to solve the ISI problem. Finally, through the comparison with existing channel equalization methods, its performance benefits of both ISI elimination and convergence speed are verified. The research findings have theoretical significance in a high-speed FSO communication system.
Laser communication has become the main driving force for the development of modern wireless optical communication technology with the characteristics of large communication capacity, good concealment and good directivity. Acquisition, pointing and tracking system which is refer to as the APT system is the key technology to the laser communication, and the detection and processing technology is one of the key technologies of APT system. The effect of scintillation and laser spot drift caused by atmospheric turbulence seriously affects the laser spot positioning accuracy of laser communication APT system, which will affect the performance of laser communication system. It is very important to choose the appropriate spot pre-processing method and the best method to improve the positioning accuracy of laser communication system. An ideal spot image with known center coordinates was generated artificially and the MATLAB was used to simulate the atmospheric turbulence to make the laser spot close to the real atmosphere. Chosing median filtering and mean filtering method to the denoising pretreatment to get filtered image. Using iterative threshold method to obtain the binary image. Through 3 common spot positioning method like the gray centroid method, circle fitting method and the Gaussian fitting method to calculate the centroid of the binary image. After getting the central coordinates, the results were compared and analyzed. Experimental results showed that mean filtering is better than median filter to filter noise of the laser spot. Compared with other methods, the centroid accuracy obtained by the gray centroid method had larger deviation due to the process of filtering the noise was not completely suppressed. The laser spot center calculating by the Gaussian fitting method were with higher positioning accuracy. According to the calculation results, the applicable conditions of different spot location algorithms were given.
Orthogonal frequency division multiplexing (OFDM) technique applied to the atmospheric optical communication can improve data transmission rate, restrain pulse interference, and reduce effect of multipath caused by atmospheric scattering. Channel estimation, as one of the important modules in OFDM, has been investigated thoroughly and widely with great progress. In atmospheric optical communication system, channel estimation methods based on pilot are common approaches, such as traditional least-squares (LS) algorithm and minimum mean square error (MMSE) algorithm. However, sensitivity of the noise effects and high complexity of computation are shortcomings of LS algorithm and MMSE algorithm, respectively. Here, a new method based on compressive sensing is proposed to estimate the channel state information of atmospheric optical communication OFDM system, especially when the condition is closely associated with turbulence. Firstly, time-varying channel model is established under the condition of turbulence. Then, in consideration of multipath effect, sparse channel model is available for compressive sensing. And, the pilot signal is reconstructed with orthogonal matching tracking (OMP) algorithm, which is used for reconstruction. By contrast, the work of channel estimation is completed by LS algorithm as well. After that, simulations are conducted respectively in two different indexes -signal error rate (SER) and mean square error (MSE). Finally, result shows that compared with LS algorithm, the application of compressive sensing can improve the performance of SER and MSE. Theoretical analysis and simulation results show that the proposed method is reasonable and efficient.