Algorithm and VLSI Implementation for Double-Binary Turbo Codes in Wimax

This paper designs a double-binary Turbo decoder applied in Wimax system. The decoder efficiently compensates the error resulted from the max calculation in the Max-log-MAP algorithm by amending the convention algorithm with only small increase in complexity. In addition, a new generalized interleaver structure which satisfies all codeword lengths in Wimax is proposed. Simulation results show that, the amending algorithm can offer approximately 0.35 to 0.4 dB gain compared with the convention algorithm at bit error rate (BER) of 10^-4. The proposed interleaver structure significantly reduces the computational complexity and areas. A double-binary turbo decoder is designed for Wimax can achieve the throughput of 2O.91 Mb／s at the operation frequency of 200 MHz, which is sufficient for the data rate of the system.