Research Of Decoding Algorithms For Turbo Code And Its Applications

Because the performance of Turbo code has achieved the shannon capacity, recently it has been improved and used in many fields. But it has a few faults, such as, the floor error, the high delay and operation in decoding . In this paper, to lower the delay and the operation is our purpose.Firstly, A few parameters influenced the performance of turbo code, such as , the rate of code, the length of inter-leaver and the number of iteration, which have been studied and discussed. And the simulations have provided the basic principle in selecting the parameter for designing of Turbo code.Secondly, based on the parallel principle of mutli-CPU, we presented a new method, that is, T-BCJR algorithm was used in a new parallel structure. Then we discussed the virtue of the T-BCJR algorithm used in the new structure.Thirdly, by studied and compared the present stopping criterions of iterative decoding, based on the notion of preference 64, we presented a new stopping criterion of iterative decoding which called CRC-SDR stopping criterion. The new criterion reduced the operation complexity and storage requirement. By simulated, the performance of Turbo code has not been degraded greatly, but the number of iteration has been decreased greatly.At last, by observing and studying the convergence of Turbo code, we presented two stopping criterions that are based on the minimum and mean of the magnitude of LLR. Simulations has been illustrated that the performance of Turbo code has not been degraded greatly and that the number of iteration has been decreased greatly.