Multi-Layer Integrated Resource Management for Mobile Wireless Systems

该奖项是根据2009年美国复苏和再投资法案（公法111-5）资助的。本研究的目的是设计和原型移动无线硬件，在广泛的性能范围内保持高能效。该方法系统地解决了多层无线硬件的低效率问题。无线硬件通常是为提供峰值性能而优化的；随着实现的性能下降，能源效率急剧降低。关键的方法是设计可伸缩的硬件，并根据瓶颈进行合理的伸缩。在知识价值方面，本项目以四个工程贡献为目标。首先，对无线通信的资源需求进行建模，设计出共同优化速率自适应和功率管理的自适应策略。其次，开发了可以根据数据速率和功率要求适当扩展的无线检测体系结构，并与无线系统中的高层集成。第三，开发了可配置的解码器架构，以细粒度的方式动态调整解码器。第四，利用Rice WARP平台，将所提出的技术实现到一个完整的功能无线硬件系统中，以评估各种无线应用下的解决方案。就更广泛的影响而言，与卫生科学家在预防保健方面的多学科合作有可能对社会需求作出贡献。收集的数据、研究评价平台和教学内容将实现开源和在线。与包括德州仪器和诺基亚在内的行业领导者的持续合作，有助于将技术转移到未来的移动无线硬件中。本研究的结果被纳入VLSI设计和移动计算的本科课程。

This award is funded under the American Recovery and Reinvestment Act of 2009 (Public Law 111-5).The objective of this research is to design and prototype mobile wireless hardware that retains high energy efficiency across a broad performance range. The approach is to systematically address the inefficiency in multiple layers of wireless hardware. Wireless hardware is conventionally optimized for delivering peak performance; the energy efficiency reduces drastically as the achieved performance decreases. The key methodology is to design scalable hardware and judiciously scale it according to the bottleneck.With respect to intellectual merit, this project targets four engineering contributions. First, the resource demands of wireless traffic are modeled to design adaptive policies that jointly optimize rate adaption and power management. Second, wireless detection architectures are developed that can scale properly based on the data rate and power requirements, and integrate with the higher levels in the wireless system. Third, configurable decoder architectures are developed that dynamically adjust the decoder in a fine-grain manner. Fourth, using the Rice WARP platform, the proposed technologies are implemented into a complete functional wireless hardware system to evaluate the solutions under a wide range of wireless applications.With respect to broader impact, the multidisciplinary collaboration with health scientists in preventive health care has the potential for contribution to societal needs. The collected data, the research evaluation platform, and educational content are to be made open-source and on-line. Ongoing collaboration with industry leaders, including Texas Instruments and Nokia, assist the transfer of technologies into future mobile wireless hardware. The results of this research are incorporated into undergraduate courses in VLSI design and mobile computing.