Implementation And Performance Of A Turbo Decoder For CCSDS Standard

In deep space communications,one of the problems is how to transmit message credibly in low signal-to-noise ratio(SNR)channel environment.To achieve reliable communication over a noisy channel,channel coding has been used since 1960s. Turbo codes are error-correcting codes based on parallel concatenation of convolutional codes and can achieve near channel capacity performance with a long block size and many decoding iterations.Due to the near-optimum performance ,Turbo codes have been an add-on option to the CCSDS telemetry channel coding recommendation without modifying the existing coding schemes and retained compatibility with the CCSDS packet telemetry recommendation.Given turbo encoder structure specified by CCSDS, this paper addresses the implementation of a simplified and efficient Turbo decoder in FPGA technology. To implement turbo decoder ,it adopted the sliding window method and devised a way to simplify max* function of Log-MAP algorithm.First,this paper presents some background knowledge of deep space communications.In the next section,it describes Turbo coding/decoding principles and two decoding algorithms:MAP and Log-MAP.After a brief overview of turbo encoder of the CCSDS standard,it analyses those factors that will influence its performance with the experimental results,and introduces how to choose some variables wisely in the implementation of Log-MAP decoder.The dedicated turbo decoder architecture and the FPGA prototype are described in the next section. Finally it gives the overall conclusions of the paper.