CAREER: Energy Scalable Signal Processing for Energy Harvesting Microsystems

申请编号：ecs - 0547113摘要该申请的重点是开发能量收集微系统的集成电路。数字技术的进步将计算带出了机房，使其几乎无处不在。将计算与传感和驱动相结合，可以在交通系统、环境研究和公共安全方面实现新的应用。然而，电池技术并没有跟上时代的步伐——限制了尺寸、使用寿命，并提高了成本。从外部资源收集能量可以实现下一代无处不在的计算，但能量收集微系统仍处于起步阶段。此外，对更小设备和高集成度的渴望限制了能量收集的可用功率。智力优势：本提案通过开发用于传感器信号处理的微架构和电路来解决这些问题，这些微架构和电路具有节能、能量可扩展和对电压变化的鲁棒性，这些都是能量收集传感器的特征。该方法是通过使用能量回收和自定时电路开发不需要直流电源的计算元件，以适应交流电源的能量收集操作。该项目最终开发了一种定制的DSP集成电路，目标能量效率为10 TeraOps/W，这将展示未来能量收集微系统的关键使能技术。更广泛的影响：为了扩大这项工作的影响，该项目将通过设计项目和实验室实验将能源可扩展性的想法引入新的和修订的课程。目标是将物理限制重新引入数字设计，目前由逻辑综合主导，并将教育与研究计划联系起来。该项目将通过美国国家科学基金会加州少数民族参与联盟（CAMP）计划，在拟议的研究中涉及代表性不足的少数民族。该项目将通过与加州大学伯克利分校的研究人员以及摩托罗拉和赛灵思公司的行业研究人员的跨学科合作，在网上传播方法、结果和工具，为能源收集和低功耗研究社区做出贡献。

CAREER: Energy Scalable Signal Processing for Energy Harvesting MicrosystemsPI: Rajeevan AmirtharajahProposal No: ECS-0547113AbstractThis CAREER proposal focuses on developing integrated circuits for energy harvesting microsystems. Advances in digital technology have brought computation out of the machine room and made it nearly ubiquitous. Integrating computation with sensing and actuation enables new applications in transportation systems, environmental studies, and public safety. However, battery technology has not kept pace - limiting size, operating lifetime, and raising costs. Energy harvesting from external sources can enable the next generation of ubiquitous computation, but energy-harvesting microsystems are still in their infancy. Moreover, the desire for smaller devices and high integration limits the power available from energy harvesting. Intellectual Merit: This proposal addresses these issues by developing microarchitectures and circuits for sensor signal processing that are energy efficient, energy scalable, and robust to voltage variations, which characterize energy harvesting transducers. The approach is to develop computational elements that do not require DC power supplies by using energy recovery and self-timed circuits adapted to energy harvesting operation from AC supplies. The project culminates in the development of a custom DSP integrated circuit with a target energy efficiency of 10 TeraOps/W, which will demonstrate critical enabling technologies for future energy harvesting microsystems.Broader Impacts: To broaden the impact of this work, the project will bring energy scalability ideas into new and revised courses through design projects and laboratory experiments. The goal is to reintroduce physical constraints into digital design, which is currently dominated by logic synthesis, and link education with the research program. The project will involve underrepresented minorities in the proposed research through the NSF California Alliance for Minority Participation (CAMP) program. The project will contribute to the energy harvesting and low power research community by disseminating methodologies, results, and tools online and through interdisciplinary collaboration with researchers at the University of California at Berkeley and with industry researchers at Motorola and Xilinx, Inc.