Research for Mixed Signal Electronic Technologies: A Joint Initiative Between NSF and SRC: Inter-Pulse-Interval Based Mixed Signal Representations

混合信号电子技术研究：NSF 和 SRC 之间的联合倡议：基于脉冲间间隔的混合信号表示

• 批准号: 0120369
• 负责人: Daniel Hammerstrom
• 金额: $ 15万
• 依托单位: Oregon Health & Science University
• 依托单位国家: 美国
• 项目类别: Continuing Grant
• 财政年份: 2001
• 资助国家: 美国
• 起止时间: 2001-10-01 至 2005-09-30
• 项目状态: 已结题
• 来源: https://www.nsf.gov/awardsearch/showAward?AWD_ID=0120369&HistoricalAwards=false
• 关键词: Research; Mixed; Signal; Electronic; Technologies

ABSTRACTAlthough digital signal representation has become almost universal, there are still many areas where an analog representation is required to interface with an analog world or to meet various other objectives such as power dissipation, frequency, or cost. In these domains analog signal representation is essential for many input modalities such as instrumentation, sensor interfaces, and communications. Likewise, there are related output applications, such as biomedical actuation and industrial control. In addition, the needs of wireless and fiber-optical communication have reinvigorated analog design. However, there are serious problems concerning how to keep these analog components on a reasonable scaling curve as Moore's law continues unabated in the digital domain, and in integrating analog representations into large, complex digital systems ("system on a chip"). The purpose of this proposal is to study a new approach to representing analog signals that we believe will integrate more cleanly into these deep submicron, single-chip systems. Today analog signals are almost exclusively represented by current or voltage quantities. Our proposal is to borrow a page from neuroscience and to use the Inter-Pulse-Interval (IPI) between single-bit, asynchronous pulses to represent analog quantities. We are proposing to develop a mixed-mode analog/digital cell library and design methodology based on IPI representation and the associated computation elements, and to engineer a case study to illustrate the outcomes. As we move to deep submicron and then on to the nanometer/molecular devices, the problems that digital encounters with scaling, such as threshold inconsistency, subthreshold currents, hot-electron effects, doping variability, substrate coupling, and transmission line and complex cross-talk effects, are even more serious for analog circuitry. IPI representations will provide significantly better immunity to these effects, as well as to the more traditional process, temperature, and reference voltage variations. For most applications, pulse based analog systems will require less power. There are numerous advantages to using pulses or pulses to communicate. They are significantly more immune to noise. An approximate analogy would be that of AM versus FM radio signal representation. The outcomes of the proposed research are Create a library of basic communication, computation (arithmetic and logic) and conversion building blocks;Design, implementation and testing of a case study using the derived building blocks and methodology;Document an IPI-based design methodology

虽然数字信号表示已经变得几乎普遍，但仍然有许多领域需要模拟表示来与模拟世界接口或满足各种其他目标，如功耗，频率或成本。在这些领域中，模拟信号表示对于许多输入方式（如仪器仪表、传感器接口和通信）是必不可少的。同样，也有相关的输出应用，如生物医学驱动和工业控制。此外，无线和光纤通信的需求使模拟设计重新焕发活力。然而，随着摩尔定律在数字领域的持续发展，以及将模拟表示集成到大型、复杂的数字系统（“片上系统”）中，如何将这些模拟元件保持在合理的缩放曲线上，存在着严重的问题。本提案的目的是研究一种表示模拟信号的新方法，我们相信这种方法将更干净地集成到这些深亚微米单芯片系统中。今天，模拟信号几乎完全由电流或电压量表示。我们的建议是借用神经科学的一页，并使用单比特异步脉冲之间的脉冲间隔（IPI）来表示模拟量。我们建议开发一个基于IPI表示和相关计算元素的混合模式模拟/数字单元库和设计方法，并设计一个案例研究来说明结果。随着我们深入亚微米，再到纳米/分子器件，数字电路遇到的缩放问题，如阈值不一致、亚阈值电流、热电子效应、掺杂可变性、衬底耦合、传输线和复杂串扰效应，对模拟电路来说更加严重。IPI表示将对这些影响以及更传统的工艺、温度和参考电压变化提供更好的免疫能力。对于大多数应用，基于脉冲的模拟系统将需要更少的功率。使用脉冲或脉冲进行通信有许多优点。他们对噪音的免疫力明显更高。一个近似的类比是调幅与调频的无线电信号表示。提出的研究成果是：创建一个基本通信、计算（算术和逻辑）和转换构建块的库；使用衍生的构建模块和方法设计、实施和测试案例研究；记录基于ipi的设计方法