标题: | 以记忆体为基础的多标准前端错误更正 A Memory Based Multi-Standard FEC Decoder Design |
作者: | 曾逸晨 Yi-Chen Tseng 李镇宜 Chen-Yi Lee 电子研究所 |
关键字: | 里德所罗门编码;卷积交错器;前端错误更正解码器;Reed-Solomon Codes;convolutional interleaver;FEC decoder |
公开日期: | 2003 |
摘要: | 前端错误更正在通讯系统是一个相当重要的功能,它主要包含搅拌器(scrambler)、 里德所罗门编码(Reed-Solomon coding)、交错器(interleaver)和回旋编码(trellis coding)。 对于效能和复杂度的考量,随着不同的应用而会有不同的设计参数。本论文提出一个高效率节省功率和面积架构的多标准前端错误更正解码器以符合不同系统的要求,而所提出的多标准前端错误解码器可以完全相容于ITU-T J.83的缆线数据系统并且可相容于数位视讯广播和ATSC数位电视等系统。我们所提出的多标准前端错误更正解码器主要包含一以记忆体来储存及更正资料的多模里德所罗门解码器和一以记忆体为基础及位址产生器的泛用型卷积解交错器(convolutional de-interleaver),皆具有以最小晶片面积达到最多功能的特色。以0.18微米 1P6M CMOS 制程实作的结果大约需要5万4千个逻辑闸以及6千位元的嵌入式静态随机存取记忆体,最快可以达到83MHz(600Mbps)的工作频率。平均消耗功率在最复杂的解码模式及在83MHz之工作频率下大约是45mW;在满足系统规范之操作参数环境下,平均消耗功率大约5.4mW。 Forward Error Correction (FEC) which mostly contains scrambler, Reed-Solomon coding, interleaving, and trellis coding is a key component in communication system. For the performance and complexity issues, design parameters are different in various applications. In this thesis, a multi-standard FEC decoder is presented to meet different system requirements with a power and area efficient architecture. The proposed multi-standard FEC decoder is fully compliant to ITU-T J.83 cable modem system and is also compatible to DVB-T and ATSC Digital TV, etc. The proposed multi-standard FEC decoder, including a multi-mode Reed-Solomon decoder with memories to store and correct the received data and a memory-based universal convolutional interleaver with a simple address generator, has the advantage of lowest overhead. With 0.18mm 1P6M CMOS technology, the implemented chip shows the FEC decoder can work at 83MHz (600Mbps) while costs 54.5K gate counts and two 376x8 bits embedded duel-port SRAM. The average power consumption in full spec. mode is about 45mW at 83Mhz. While running at 7MHz that meets symbol rate of cable modem, the power dissipation is 5.4mW. |
URI: | http://140.113.39.130/cdrfb3/record/nctu/#GT009111585 http://hdl.handle.net/11536/43491 |
显示于类别: | Thesis |
文件中的档案:
If it is a zip file, please download the file and unzip it, then open index.html in a browser to view the full text content.