High-Throughput Reconfigurable Convolutional Turbo Decoder Design for Future Wireless WAN
Date Issued
2009
Date
2009
Author(s)
Chen, Chun-Yu
Abstract
With the rapid growth of multimedia service demand, the data-rate requirement will be higher than 100 Mbps for future wireless WAN, such as 3GPP-LTE and WiMAX systems. In recent years, these wireless WAN systems have adopted advanced channel coding schemes in order to increase transmission reliability. Among them, the single-binary convolutional turbo code (SB-CTC) introduced by Berrou et al. in 1993 had been proved to have decoding performance close to the Shannon limit. Then, in 1999, the double-binary CTC (DB-CTC) was proposed to encode/decode two bits per time and yield to better coding gain compared with the SB-CTC. The SB- and DB-CTC schemes have been adopted in 3GPP-LTE and WiMAX systems, respectively.he goal of this thesis is to design a high-throughput reconfigurable CTC decoder that can be used in future wireless WAN systems. The features are 1) propose new architecture for the MAP decoder and interleaver, the two main components of the CTC decoder, to achieve high-throughput requirement, 2) design the reconfigurable MAP decoder and interleaver based on the computational similarity between SB-CTC and DB-CTC decoding algorithms, and 3) target the CTC specification of 3GPP-LTE and WiMAX systems. The synthesis results show that the hardware overhead for the MAP decoder to support both CTC decoding algorithms is less than 10%. And only 64.2% hardware resource is required to generate the interleaved address of both 3GPP-LTE and WiMAX systems.inally, the proposed MAP decoder and interleaver are embedded into a CTC decoder. By using TSMC 0.13 μm 1P8M CMOS process, the proposed CTC decoder is implemented in a chip at 147 MHz maximum operating frequency with core size of 7.02 mm2. The implemented decoder can support 34-mode CTC decoding for both 3GPP-LTE and WiMAX systems. The maximum throughput rate can achieve 260.2 Mbps which satisfies the data-rate requirement of the future wireless WAN systems.
Subjects
Convolutional Turbo Code
high-throughput
reconfigurable
parallel
Type
thesis
File(s)![Thumbnail Image]()
Loading...
Name
ntu-98-R95943173-1.pdf
Size
23.32 KB
Format
Adobe PDF
Checksum
(MD5):c9ae8956a83d0c3399e39139b53f830f