{"title":"High throughput hardware architecture for (1440,1344) low-density parity-check code utilizing quasi-cyclic structure","authors":"H. Yamagishi, M. Noda","doi":"10.1109/TURBOCODING.2008.4658676","DOIUrl":null,"url":null,"abstract":"High throughput architecture of an encoder and a decoder for a quasi-cyclic low-density parity-check (LDPC) code is proposed. A new systematic encoding method is carried out by polynomial manipulation. The proposed decoder architecture, where the check-node process is split into two processes so that the memory access becomes column-wise, enables overlapped message-passing for any parity-check matrix. The hardware architecture for the check-node processes utilizing a quasi-cyclic structure does not require complex multiplexers. Hardware employing the proposed architecture for a (1440,1344) LDPC code designed for high throughput millimeter wave application is evaluated using 65 nm CMOS technology. The gate count of the encoder for 3 Gbps and 6 Gbps throughput is 2.5 k and 3.1 k, respectively, and the gate count of the decoder for 8 iterations is 304 k and 409 k, respectively. A bit-error rate of 10-6 is obtained at Eb/N0 of 5.9 dB, and the estimated power consumption of the decoder is 58 mW for 3 Gbps and 86 mW for 6 Gbps.","PeriodicalId":284046,"journal":{"name":"2008 5th International Symposium on Turbo Codes and Related Topics","volume":"1 1","pages":"0"},"PeriodicalIF":0.0000,"publicationDate":"2008-10-24","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":"18","resultStr":null,"platform":"Semanticscholar","paperid":null,"PeriodicalName":"2008 5th International Symposium on Turbo Codes and Related Topics","FirstCategoryId":"1085","ListUrlMain":"https://doi.org/10.1109/TURBOCODING.2008.4658676","RegionNum":0,"RegionCategory":null,"ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":null,"EPubDate":"","PubModel":"","JCR":"","JCRName":"","Score":null,"Total":0}
引用次数: 18
Abstract
High throughput architecture of an encoder and a decoder for a quasi-cyclic low-density parity-check (LDPC) code is proposed. A new systematic encoding method is carried out by polynomial manipulation. The proposed decoder architecture, where the check-node process is split into two processes so that the memory access becomes column-wise, enables overlapped message-passing for any parity-check matrix. The hardware architecture for the check-node processes utilizing a quasi-cyclic structure does not require complex multiplexers. Hardware employing the proposed architecture for a (1440,1344) LDPC code designed for high throughput millimeter wave application is evaluated using 65 nm CMOS technology. The gate count of the encoder for 3 Gbps and 6 Gbps throughput is 2.5 k and 3.1 k, respectively, and the gate count of the decoder for 8 iterations is 304 k and 409 k, respectively. A bit-error rate of 10-6 is obtained at Eb/N0 of 5.9 dB, and the estimated power consumption of the decoder is 58 mW for 3 Gbps and 86 mW for 6 Gbps.