Research And Implementation Of Channel Estimation And Diversity Reception In DVB-T2 System

Digital video terrestrial broadcasting is the most widely used and basic form of transmission in digital television broadcasting,but at the same time,its transmission conditions are the worst and the interference is the most serious.This paper mainly studies the second generation digital video broadcasting-terrestrial(DVB-T2)system.Aiming at the problem of poor reception performance in high-speed mobile application environment,two research directions of this paper are pointed out: channel estimation and diversity merging algorithm in dynamic multipath fading channel.Firstly,the propagation characteristics of wireless channel,the formation of time-frequency double-selective fading channel,the frame structure of DVB-T2 system and the type of pilot are introduced in detail.Aiming at the insertion mode of discrete pilot in DVB-T2,the channel estimation algorithm of pilot-based orthogonal frequency division multiplexing(OFDM)system is introduced.The least square(LS)estimation algorithm,linear least mean square error(LMMSE)estimation algorithm and two-dimensional Wiener are studied.Interpolation algorithm,time-frequency weighted interpolation algorithm and so on.At the same time,the performance of the above channel estimation algorithm in dynamic multipath fading channel is simulated and analyzed.Due to the introduction of multiple input single output(MISO)technology in DVB-T2,aiming at the particularity of discrete pilot in DVBT2 MISO mode,this paper introduces the channel estimation algorithm adapting to its system,and proposes an improved algorithm in this paper.The adaptive time-frequency weighted channel estimation has better performance for time-varying channel estimation without increasing the number of pilots.Subsequently,the signal detection algorithm in DVB-T2 MISO mode is studied.It is concluded that zero forcing(ZF)detection is simple,but its performance is poor.The minimum mean square error(MMSE)detection algorithm balances interference rejection and noise amplification,and its performance is slightly better than zero forcing algorithm.Maximum likelihood(ML)detection has the best performance,but its complexity increases with the number of antennas.Then,the MISO mode of DVB-T2 is extended to multiple input multiple output(MIMO)mode.Different merging strategies under two receiving antennas are introduced,and the complexity and performance of each strategy are compared.Comparisons show that the maximal ratio combining(MRC)performance is the best,but the complexity is also the highest.Finally,for DVB-T2 system,the implementation and verification of channel estimation and aggregation based on FPGA are carried out.Through the above research and discussion,the performance and feasibility of the balancing algorithm are balanced,and the final hardware implementation scheme is determined.LS algorithm and time-frequency linear interpolation are used in channel estimation,and the maximum ratio combination of received signals of branch is carried out.On the platform of Quartus II,it is implemented in Verilog language,and the correctness of the designed circuit is verified.