Research On Channel Estimation Techniques For Wireless Optical OFDM Communication Systems Based On Compressed Sensing

Optical orthogonal frequency division multiplexing(OFDM)is one of the important research directions of wireless optical communication technology at present.In wireless optical communication systems,channel estimation acts as the premise of signal recovery,equalization and other technologies,which is critical to ensuring the reliability of communication systems.Owing to the differences in system between traditional OFDM system and asymmetrically clipped optical orthogonal frequency division multiplexing(ACO-OFDM)sytem,many channel estimation methods are not applicable for ACO-OFDM system.In recent years,the application of compressed sensing theory in channel estimation has been widely studied,which exerts positive functions in the improvement of estimation accuracy and system performance.With the growing attention on wireless optical OFDM communication technology,the study of wireless optical OFDM system channel estimation technology based on compressed sensing developes significant meaning for the guarantee of the reliability of wireless optical communication system.In this thesis,the pilots based channel estimation methods are analyzed,aiming at the problems which include the lack of estimation accuracy and low subcarrier utilization efficiency,the compressed sensing based channel estimation method for ACO-OFDM systems is proposed.According to the characteristics of ACO-OFDM system,the pilot allocation optimization algorithm for ACO-OFDM system is proposed.Firstly,the pilot symbols with Hermitian symmetrical structure were inserted in odd subcarriers to satisfy the requirements of ACO-OFDM.Then,the pilot allocation was optimized by minimizing the summation of squared columns correlation of the measurement matrix.Finally,the orthogonal matching pursuit(OMP)algorithm is used to estimate channel response coefficient.In comparison with the pilot based channel estimation,the channel estimation based on compressed sensing uses fewer pilot symbols,which improves the subcarriers utilization efficiency of ACO-OFDM system.Moreover,the proposed pilot allocation optimization algorithm is more feasible than the proof of restricted isometry property(RIP),thus possessing higher practical values.Secondly,aiming at the defects that OMP algorithm can not adapt to sparsity and lack of estimation accuracy,a modified sparsity adaptive-generalized orthogonal matching pursuit algorithm(MSA-GOMP)is proposed.This algorithm estimates the sparsity with sparsity adaptive matching pursuit(SAMP)algorithm and improves the shortcomings caused by fixed step size.The secondary estimation will be conducted with generalized orthogonal matching pursuit(GOMP)algorithm in order to improve channel estimation accuracy.Comparing with OMP,the proposed channel estimation algorithm could adapt to channel sparsity and improve channel estimation accuracy,which has certain significance to improve ACO-OFDM channel estimation accuracy and decrease bit error rate.At last,the research content of this thesis is summarized and future research directions are given.