A Low-Power Variable Length Decoder for MPEG-2 with Self Checking FIFO

碩士 === 國立彰化師範大學 === 資訊工程學系 === 99 === Variable-Length Code Decoder is a widely used technique in digital video compression systems. The length of each decoding are not fixed in Variable length decoder, so if an data error occurred before the data transformed into the variable length decoding, this e...

Full description

Bibliographic Details
Main Author: 葉皇輝
Other Authors: 陳仁德
Format: Others
Language:zh-TW
Published: 2011
Online Access:http://ndltd.ncl.edu.tw/handle/31099351967192264156
Description
Summary:碩士 === 國立彰化師範大學 === 資訊工程學系 === 99 === Variable-Length Code Decoder is a widely used technique in digital video compression systems. The length of each decoding are not fixed in Variable length decoder, so if an data error occurred before the data transformed into the variable length decoding, this error data will lead to the next data error. Using Variable-Length Code Decoder can reach high compression efficiency, but how to solve this error data will be a big problem. In order to solve the problem of error data, there are some methods in MPEG standard, these methods are focused on how to give up these error data. The less data discard, the more accurate decode will be. In this paper we propose a method, we use Self Checking FIFO Queue in Variable-Length Code Decoder, this method can solve the intermittent error, and can detect the error data earlier, and send the error sign to Variable-Length Code Decoder, to avoid error decoding, use this method can get better decoding efficiency. In original Self Checking FIFO Queue, it uses Berger code to generate check bit, and than uses check bit to check the data correctness. But using Berger code to generate check bit, this method cannot detect the error data in some conditions. So in this paper, we improve the method of Berger code to generate check bit. We called the new methods “Double Berger code”, this new method improved the check bit generating method of Berger code and greatly increase the error data detection rate.