CSR:Small: Novel Techniques for Lossless Data Compression and Efficient Decompression in Heterogeneous Embedded Systems

CSR：Small：异构嵌入式系统中无损数据压缩和高效解压缩的新技术

• 批准号: 0915376
• 负责人: Prabhat Mishra
• 金额: $ 10万
• 依托单位: University of Florida
• 依托单位国家: 美国
• 项目类别: Standard Grant
• 财政年份: 2009
• 资助国家: 美国
• 起止时间: 2009-08-01 至 2012-07-31
• 项目状态: 已结题
• 来源: https://www.nsf.gov/awardsearch/showAward?AWD_ID=0915376&HistoricalAwards=false
• 关键词: CSR; Small; Novel; Techniques; Lossless

Abstract0915376 - Novel Techniques for Lossless Data Compression and Efficient Decompression in Heterogeneous Embedded SystemsPI: Prabhat MishraDemands for heterogeneous and complex embedded applications have soared drastically in recent years. Memory is one of the key driving factors in designing such systems since a larger memory indicates an increased chip area, more power dissipation, and higher cost. Compression techniques are promising to reduce memory requirements by reducing the program and data size. The existing approaches either perform efficient compression using complex and variable-length encodings at the cost of slow decompression, or fast decompression using simple and fixed-length encodings at the cost of compression efficiency. This research will develop novel tools and techniques to achieve both efficient compression and fast decompression for a wide variety of application domains and system architectures. This project will make three fundamental contributions: novel decode-aware compression, optimal bitstream placement for parallel decompression, and synergistic integration of compression and encryption. A successful implementation of the proposed research will result in significant improvement in system performance and security, and drastic reduction in overall area, cost and energy requirements. Synergistic integration of compression and encryption will lead to efficient and secure embedded systems. The outcome of this research has a direct impact on everyday life. Improved compression/decompression techniques will have double impact ? low-cost and low-power everyday appliances for the public and improved performance and security for safety-critical systems. This project will have broad educational impact by developing compression related projects in both graduate and undergraduate courses, and by involving undergraduates as well as minority students through UF Honors, University Minority Mentoring and SEAGEP programs.

Abstract0915376-近年来，异质嵌入的SystemSPI中无损数据压缩和有效减压的新技术：用于异质和复杂的嵌入式应用程序的Prabhat Mishrademands近年来已经猛增了。记忆是设计此类系统的关键驱动因素之一，因为更大的内存表明芯片区域增加，更多的功率耗散和更高的成本。压缩技术有望通过减少程序和数据大小来减少内存需求。现有方法要么以缓慢的减压成本使用复杂和可变的长度编码执行有效的压缩，要么使用简单和固定长度的编码以压缩效率为代价进行快速减压。 这项研究将开发新的工具和技术，以实现各种应用领域和系统体系结构的有效压缩和快速减压。该项目将做出三个基本贡献：新颖的解座感知压缩，对并行解压缩的最佳位置放置以及压缩和加密的协同整合。拟议研究的成功实施将导致系统性能和安全性的显着提高，并大大降低了整体领域，成本和能源需求。压缩和加密的协同整合将导致有效且安全的嵌入式系统。这项研究的结果对日常生活有直接影响。改进的压缩/减压技术会产生双重影响吗？低成本和低功耗的日常设备，用于公众，改善了安全至关重要系统的性能和安全性。该项目将通过在研究生和本科课程中开发与压缩相关的项目，以及通过UF荣誉，大学少数族裔指导和SEAGEP计划参与本科生以及少数族裔学生，从而产生广泛的教育影响。