Research And Realization On DSP About The Decoding Of Convolutional Code

From P.Elias first gave the concept of convolutional code, it has show its'great advantage. Under the same condition and the same rate of block code, the performance of convolutional code is better than block code, and it's easier to implement the best decoding.Now, convolutional code has been widely applied in wireless communication standards, such as GSM, CDMA2000 and IS-95 .etc.With the in-depth study of channel coding theory,the people have developd new kinds of convolutional code. Block convolutional code and Turbo code are two typical types. In addition, there are Woven convolutional code and Punctured convolutional code etc. In the future, the wireless communication system will need higher performance to meet the development of communication business, so it will lead the appearance of more codes.This paper introduce the principle of coding and decoding,and explan the characteristic,error correction ability,performance boundary and coding gain of convolutional code. Analyse and compare the decoding performance under different rate,different constraint length and different decision way,at last get the conclusion.When we use the high rate code rate, the system will be more complex.In order to solve the problem, we can use punctured convolutional code to get high rate code through making use of simple convolutional code.By simulation, we can see the performance of punctured convolutional code is better.At last, this paper achieves the design of the decoder bases on DSP in digital transmitter-receiver. Referring to the thought of the complexity of coding and the ability of error-correction, we choose the NASA convolutional codes of constraint length 7, rate 1/2 and select TMS320C5402 for the DSP device. Finally, the paper gives the procedure. Run on the CCS, the result of coding and encoding is right, and we can confirm the error-correcting capacity of (2,1,7) convolutional code.