Development of a Functional LSI Achieving Low-rate Multimedia Data Transmission.

开发实现低速率多媒体数据传输的功能LSI。

• 批准号: 10450136
• 负责人: ONODERA Hidetoshi
• 金额: $ 6.98万
• 依托单位: KYOTO UNIVERSITY
• 依托单位国家: 日本
• 项目类别: Grant-in-Aid for Scientific Research (B).
• 财政年份: 1998
• 资助国家: 日本
• 起止时间: 1998 至 2000
• 项目状态: 已结题
• 来源: https://kaken.nii.ac.jp/en/grant/KAKENHI-PROJECT-10450136/
• 关键词: DSP; Multimedia; Parallel Processing; Image Compression; Low-rate; VLSI; FMPP; 機能メモリ; 高速バス

The main goal of this research project is to develop a multimedia communication system including low-power system-LSIs that are developed for real-time low-rate image compression and decompression.This research develops two innovative results as follows.(1) A Low-rate Video Compression Algorithm using Vector-Quantization : FRMSHVQ FRMSHVQ stands for fixed-rate multistage Hierarchical Vector Quantization. It can compress a 10fps. Q-CIF (176×144) video sequence to 29.2kbps. FRMSHVQ uses vector quantization instead of DCT on MPEG which feature enables easy decompression by using a table look-up compared with IDCT operation on MPEG.(2) A Low-power DSP for FRMSHVQ : VP-DSP We have developed and implemented a low-power DSP for the FRMSHVQ algorithm called VP-DSP.VP-DSP stands for a vector-pipeline DSP.The register file of VP-DSP contains several long-bit (160-bit) registers for 16-parallel SIMD operations frequently used on FRMSHVQ.The implemented LSI works properly at 25MHz and 1.6V.The power dissipation is only 49mW for real-time encoding/decoding of a 10fps. QCIF video sequence.

本研究项目的主要目标是开发一种包含低功耗系统的多媒体通信系统-LSI，用于实时低码率图像压缩和解压缩。本研究取得了以下两个创新性成果：(1)一种基于矢量量化的低码率视频压缩算法：FRMSHVQ FRMSHVQ代表固定速率多级分层矢量量化。它可以压缩10fps。Q-CIF(176×144)视频序列到29.2kbps。(2)FRMSHVQ的低功耗数字信号处理器(VP-DSP)我们为FRMSHVQ算法开发并实现了一种低功耗的数字信号处理器VP-DSP。VP-DSP是一种矢量流水线的数字信号处理器。VP-DSP的寄存器文件中包含多个长位(160位)的寄存器，用于FRMSHVQ上常用的16位并行SIMD运算。所实现的LSI在25 MHz和1.6V下工作正常。对于10fps的实时编/解码，功耗仅为49 mW。QCIF视频序列。

项目成果

T.Iwahashi, T.Shibayama, M.Hahsimoto, K.Kobayashi, H.Onodera: "Vector Quantization Processor for Mobile Video Communication."Proc.of 13th IEEE SOC/ASIC Conference. 2000. 75-79 (2000)

T.Iwahashi、T.Shibayama、M.Hahsimoto、K.Kobayashi、H.Onodera：“用于移动视频通信的矢量量化处理器”。第 13 届 IEEE SOC/ASIC 会议的会议记录。

H.Onodera,K.Kobayashi: "Vector Quantization Processor for Mobile Video Communication."Proc.of 13th IEEE SOC/ASIC Conference 2000. 75-79 (2000)

H.Onodera,K.Kobayashi：“用于移动视频通信的矢量量化处理器。”Proc.of 13th IEEE SOC/ASIC Conference 2000. 75-79 (2000)

H.Onodera,K.Kobayashi: "Architecture and Performance Evaluation of a New Functional Memory : Functioal Memory for Addition"IEICE Transactions Fundamentals. E83-A(12). 2400-2408 (2000)

H.Onodera、K.Kobayashi：“新功能存储器的架构和性能评估：用于加法的功能存储器”IEICE 交易基础知识。

H.Onodera,K.Kobayashi: "Architecture and Performance Evaluation of a New Functional Memory : Functional Memory for Addition"IEICE Transactions Fundamentals. E83-A(12). 2400-2408 (2000)

H.Onodera、K.Kobayashi：“新功能存储器的架构和性能评估：加法功能存储器”IEICE Transactions Fundamentals。

K.Kobayashi: "A Real-Time Low-Rate Video Compression Algorithm Using Multi-Stage Hierachical Vector Quantization"IEICE Trans.on Fundemental. E82-A(2). 215-222 (1999)

K.Kobayashi：“使用多级分层矢量量化的实时低速率视频压缩算法”IEICE Trans.on Fundational。