Research And Implementation Of High-data-rate Turbo Encoding And Decoding For B3G Mobile Communication System

With the development of personal communication and the requirements for high-data-rate transmission in mobile system, more attentions are paid into the research of the Next-Generation Mobile Communications (beyond 3G, B3G). Professor C. Berrou proposed Turbo code in the ICC'93, which can almost reach the Shannon limitation, with only 0.7dB away. The idea of random coding was implemented by the combination of convolutional code and random interleaver, meanwhile the SISO (Soft Input Soft Output) iterative decoding was used to approach maximum likelihood decoding. Due to its own characteristic and excellent performance, Turbo code had already been applied in the 3rd generation mobile communication systems.In the B3G mobile communication systems, the throughput is even higher, which approachs 100Mbps. In this reason, Turbo code is hard to be implemented because of its iterative decoding algorithm. We must use some parallel decoding blocks to achieve to requirement of the throughput. In this paper, we first researched the algorithm of coding and decoding for Turbo code. Then, we suggested a simplified algorithm for FPGA implementation based on the Log-MAP decoding algorithm. In our algorithm, for one hand, we divided the decoding block into many sub-blocks, which saved the hardware resources. For another hand we doubled the resources of decoding blocks to meet the requirement of 100Mbps, which was the design skill in FPGA. As a result, the coding and decoding blocks were implemented and were used in the uplink of B3G-TDD system. Test under the wireless environment shows the performance and delay of our coding and decoding blocks is suitable for the requirement of B3G mobile communication system.