English  |  正體中文  |  简体中文  |  Items with full text/Total items : 75369/75369 (100%)
Visitors : 25471108      Online Users : 178
RC Version 7.0 © Powered By DSPACE, MIT. Enhanced by NTU Library IR team.
Scope Tips:
  • please add "double quotation mark" for query phrases to get precise results
  • please goto advance search for comprehansive author search
  • Adv. Search
    HomeLoginUploadHelpAboutAdminister Goto mobile version

    Please use this identifier to cite or link to this item: http://ir.lib.ncu.edu.tw/handle/987654321/69006

    Title: 盲源分離晶片設計之研究;A Study on Chip Design for Blind Source Separation
    Authors: 石敏;Shih,Min
    Contributors: 資訊工程學系
    Keywords: Blind Source Separation;Chip
    Date: 2015-08-31
    Issue Date: 2015-09-23 14:52:48 (UTC+8)
    Publisher: 國立中央大學
    Abstract: 盲訊號源分離在眾多領域都是很重要的議題,舉凡語音訊號處理、生物訊號處理、腦科學等,皆對此有研究的必要.在此論文中,我們探討語音上的盲訊號源分離,此分離方法將在未知混合過程及聲音源的情況下,僅憑藉由麥克風接收到的混合訊號,達成盲訊號源分離的任務.在本論文中,我們針對經過旋積混合模型的混合訊號做分離,此混合模型較接近真實世界環境.因此,針對旋積混合訊號的盲訊號分離研究是具有價值的.我們將介紹基於資訊最大化 (Informationn-Maximization, Infomax)與基於時頻聚類的盲訊號分離演算法.基於資訊最大化的盲訊號源分離系統透過更新濾波器的係數,來達到共同資訊最小化的效果.基於時頻聚類的盲訊號分離系統則在時頻域上先萃取混合訊號的特徵,再利用聚類的方式重建源訊號.然而,解混合運算量龐大,所以將此分離系統實現在超大型積體電路上是必要的.在此論文中,我們提出兩個分離系統的硬體架構,一個是基於資訊最大化的盲訊號源分離架構,另一個則是基於時頻聚類的盲訊號源分離架構.基於資訊最大化的盲訊號源分離架構主要由Infomax濾波器模組、縮放係數模組所組成.而基於時頻聚類的盲訊號源分離架構則主要由特徵擷取模組及時頻遮罩估計模組所組成.晶片的電路設計實現於TSMC 0.18μm CMOS technology,而兩者架構的整體晶片面積分別約為0.99×0.99 mm2 和1.68×1.68 mm2.;Blind source separation (BSS) is an important problem in many research fields, for example, biomedical signal processing, multimedia signal processing, and brain science. It is also a nice topic for speech signal processing. In this paper, we introduce Infomax based and TF masking based blind source separation. Infomax based BSS system minimizes the mutual information by learning rules for filter coefficients, while TF masking based BSS system separates the mixed signals by clustering the extracted features on time-frequency domain, and solves the source signals by TF masking. These two kinds of BSS methods solve the problems for convolutive mixed signals, which is more close to the environment in real world. However, the computation load used in convolutive blind source separation is so heavy that it is necessary to realize this part with Very-large-scale integration (VLSI).
    In this thesis, we proposed hardware architectures for both Infomax based BSS and TF masking based BSS. The architecture of Infomax based BSS system is composed of Infomax filtering modules, scaling factor computation modules, and a d-term module. On the other hand, the architecture of TF masking based BSS system consists of a feature extraction module and a K-Means module. Both of proposed BSS chips are implemented with TSMC 0.18μm CMOS technology, the die sizes of the proposed chips are approximately 0.99×0.99 mm2 and 1.68×1.68 mm2.
    Appears in Collections:[資訊工程研究所] 博碩士論文

    Files in This Item:

    File Description SizeFormat

    All items in NCUIR are protected by copyright, with all rights reserved.

    社群 sharing

    ::: Copyright National Central University. | 國立中央大學圖書館版權所有 | 收藏本站 | 設為首頁 | 最佳瀏覽畫面: 1024*768 | 建站日期:8-24-2009 :::
    DSpace Software Copyright © 2002-2004  MIT &  Hewlett-Packard  /   Enhanced by   NTU Library IR team Copyright ©   - 隱私權政策聲明