Research On Low-Complexity Timing Synchronization Algorithm And Performance Optimization For DCO-OFDM Systems

Orthogonal Frequency Division Multiplexing(OFDM)technology can effectively solve the problem of frequency selective fading introduced by multi-path effects in the wireless environment and high efficiency of spectrum utilization.It has been widely used in traditional wireless and wired communication systems.With the development of Internet technology,such as big data and cloud computing,the data transmission rate of short-range communication systems(such as access networks and data centers)has increased dramatically.Traditional binary-based on-off modulation method,due to its low spectrum utilization and poor dispersion resistance,requires the use of wavelength division multiplexing and complex dispersion compensation methods to increase the system's transmission capacity,and the single-wavelength transmission rate is not promoted.Optical OFDM technology take the advantages of fiber dispersion and high spectral efficiency which can effectively resist.It has been considered as one of the key technologies for realizing high-speed optical communication systems in the future.It has received great attention from the academic and industrial communities.This dissertation focuses on DCO-OFDM system,and uses numerical simulation and off-line experiment methods to research the key algorithms of digital signal processing(DSP)and transmission performance optimization algorithms.The main research contents are summarized as follows:1.According to the different FFT size in DCO-OFDM system,a timing synchronization algorithm based on the training sequence of the low complexity is researched by using the numerical simulation method and the off-line experiment method.The timing synchronization method based on training sequence cross-correlation and maximum value search method can obtain accurate timing synchronization.However,it requires a large number of multiplication operations in the implementation process of high-speed optical fiber communication systems and the complexity is very high.Combining the training sequence with the threshold decision method,cannot only reduce the accomplished complexity of the timing synchronization algorithm but also have better inter-relationship.Numerical simulation results show that the training sequence based on large FFT size can resist the influence of noise and sampling phase deviation well,and it has better synchronization performance.2.The channel estimation algorithm based on Least Squares(LS)combined with Intra-Symbol Average(ISFA)technique was studied,and the optimal number of ISFA taps at different received optical powers was determined.The LS-based channel.estimation method can accomplished simply.The channel estimation accuracy is affected by the system noise easily.Selecting the optimal taps in the ISFA technique can suppress noise interference and provide accurate channel estimation.The Experiment results show that has better EVM performance which the optimal taps of the ISFA gradually increases as the received optical power decreases,and the DCO-OFDM system based on large FFT size.3.An orthogonal cyclic matrix transform(OCT)precoding combined with a digital limiting algorithm is proposed to optimize the performance of the DCO-OFDM system.The experimental results show that he uneven frequency response of optical/electrical devices results in a large difference in signal-to-noise ratio(SNR)between different subcarriers of an OFDM signal,and a higher BER on data subcarriers with lower SNR,which affects the overall system performance.The OCT precoding technique can effectively average the SNR on each subcarrier,improving the BER performance of the system,the OCT precoded signal also has a higher peak-to-average power ratio(PAPR).In order to reduce the signal PAPR,clipping is used.Experiments show that the joint algorithm can further improve the BER performance of the system.