首页> 外文期刊>Journal of signal processing systems for signal, image, and video technology >The Design and Implementation of a Highly Efficient and Low-Complexity Joint-MMSE GFDM Receiver
【24h】

The Design and Implementation of a Highly Efficient and Low-Complexity Joint-MMSE GFDM Receiver

机译:高效且低复杂性关节MMSE GFDM接收器的设计与实现

获取原文
获取原文并翻译 | 示例
获取外文期刊封面目录资料

摘要

This paper presents the VLSI architecture of a highly-efficient and low-complexity Generalized Frequency Division Multiplexing (GFDM) receiver based on the joint Minimum Mean Square Error (joint-MMSE) approach. To be specific, a novel data processing flow is proposed in this paper where the bit reverse permutation after the FFT operation and the shuffle permutation are conducted concurrently. As a result, the processing delay is greatly reduced and the memory requirement for the temporary storage is saved. Based on the designed data flow, the VLSI architecture of the joint-MMSE GFDM receiver is presented. A novel folded structure is proposed so that the arithmetic and storage elements are utilized in a time-sharing manner and the area complexity is decreased. These elements are also architected with high configurability to support various processing modes. Moreover, a novel memory-access algorithm and address-generation circuit is designed to efficiently access the memory module and to mitigate the data conflict when supporting different processing modes. The proposed design is synthesized using TSMC 90 nm technology at operating frequency of 200 MHz. The post-synthesis estimation shows that this architecture occupies an area complexity of 492.6 kGEs. The analyses results illustrate that the proposed data flow leads to a 16% reduction in delay and complexity compared to the conventional approaches. Furthermore, the proposed design enhances the Efficiency by 4.7x compared to the state of the art work.
机译:本文介绍了基于关节最小均方误差(关节MMSE)方法的高效和低复杂性广义频分复用(GFDM)接收器的VLSI架构。具体而言,在本文中提出了一种新的数据处理流程,其中FFT操作之后的比特反向置换并同时进行。结果,大大减少了处理延迟,并保存了临时存储器的存储器要求。基于设计的数据流,提出了联合MMSE GFDM接收器的VLSI架构。提出了一种新颖的折叠结构,使得算术和存储元件以时间共享方式利用,并且面积复杂度降低。这些元素还以高配置性归档,以支持各种处理模式。此外,新的存储器访问算法和地址生成电路设计成有效地访问存储器模块并在支持不同的处理模式时减轻数据冲突。所提出的设计在200MHz的工作频率下使用TSMC 90nm技术合成。后合成后估计表明,该架构占据492.6桶的区域复杂性。分析结果说明,与传统方法相比,所提出的数据流程导致延迟和复杂度降低16%。此外,与最先进的工作相比,所提出的设计通过4.7倍提高了效率。

著录项

相似文献

  • 外文文献
  • 中文文献
  • 专利
获取原文

客服邮箱:kefu@zhangqiaokeyan.com

京公网安备:11010802029741号 ICP备案号:京ICP备15016152号-6 六维联合信息科技 (北京) 有限公司©版权所有
  • 客服微信

  • 服务号