GOALI: Physically Realizable Magnetic Information Systems

本文从介质、记录磁头和记录策略对系统性能的影响方面研究了磁性信息系统的集成问题。具体地说，写入场梯度和上升时间效应、信号稳定性、退磁和热效应、新型介质设计和探针技术将被检查。系统性能将通过系统容量、信噪比、随时间的信号退化以及对最新数据编码策略的影响来量化。提出的工作包括：量化应用领域、时间和温度对读取信号的影响；扩展现有系统模拟；发展实验技术来估计我们的介质模型中的参数；开发、模拟和实验验证与系统性能和头部模型相关的模型。这项工作涉及实验研究、模拟和理论研究之间的独特相互作用，以评估组件变化对系统性能的影响。这项广泛的研究计划建立在多年的研究努力和研究人员力量的结合之上。华盛顿大学对磁信息处理的研究一直是对磁记录系统部件的物理特性的详细研究，结合将磁信息存储和检索视为一种通信系统的信息理论分析，以及基于模型的系统性能评估。这项工作，连同对新出现的结果、技术和工具的认识，产生了对系统组件的物理特性对信息处理能力施加的限制的见解。这些知识可以指出不受目前限制的新办法，并有助于评估它们的希望和前景。随着研究人员不断提出进化和全新的磁记录系统设计，在众多选择和方向中，对多功能和可靠指南的实用性和需求是显而易见的。该提案建立在NSF之前通过NCR和ECS计划资助的研究基础上，并由NSIC， ARPA，商务部（ATP）和工业界资助，将磁信息系统的系统导向研究计划整合在一起。这项工作的核心是测量、建模和分析由介质的磁性微观结构产生的波形。这种微观结构有效地处理了写入信号，新系统的合成应该部分地由这种处理来指导。介质磁化具有峰限、非线性、滞回、粒状等特点。保留记录信息的能力最终也限制了媒介的潜在性能。系统性能对各种基本介质参数（如交换耦合、各向异性能量、粒度和薄膜厚度）的依赖性将被追求。使用更强大的分析工具来模拟这些介质效应，设计用于补偿可重复介质噪声效应的系统的性能将被量化。使用华盛顿大学磁力显微镜和记录测试仪开发的独特实验，量化了由于施加磁场、时间和温度引起的信号退化。一种基于物理模拟的磁记录系统，包括介质、读磁头和写磁头。这个计算模型在理论和实验研究之间架起了一座桥梁。它的许多特征已经通过与VSM和自旋台测量的模拟比较得到验证，包括一些相当详细的记录程序。该模型现在已达到一个发展阶段，它被设想为系统综合的工具。随着工具的改进，它必须保持模块化，以便新的头部设计，新媒体和新的信号处理模块可以很容易地合并，并评估它们的相互作用。它必须能够模拟媒体的单一实现和相应的波形，并运行蒙特卡罗模拟来预测各种性能指标的系统级性能。该模型将扩展：预测信号依赖于应用领域，时间和温度；包括当前用于磁记录信道的通信信号处理技术；包括垂直的和有图案的媒体模型；并结合探针的模型，如磁力显微镜

O'Sullivan9900159Integration issues for magnetic information systems are studied in terms of the impact of media, recording heads, and recording strategies on system performance. Specifically, write-field gradient and rise-time effects, signal stability, demagnetizing and thermal effects, novel medium design, and probe technology will be examined. System performance will be quantified through system capacity, signal to noise ratio, signal degradation over time, and impact on state-of-the-art data encoding strategies. The proposed work includes: quantifying the impact of applied fields, time, and temperature on read signals; expanding existing system simulations; developing experimental techniques to estimate parameters in our medium model; developing, simulating, and experimentally verifying models that relate system performance to medium and head models. The work involves a unique interplay between experimental studies, simulations, and theoretical studies to assess impacts of component changes on system performance. This broad research plan builds on years of research effort and the combination of the investigators' strengths.Research at Washington University on magnetic information processing has been a detailed study of the physics of magnetic recording system components, coupled with information theoretic analysis that views magnetic information storage and retrieval as a communications system, and with model-based evaluation of system performance. This work, together with cognizance of new results, techniques, and tools as they appear, yields insights into the limits imposed on information processing capacity by the physics of system components. Such knowledge can point to new approaches not subject to present limits, and serves to assess their promise and prospects. As researchers continue to propose both evolutionary and radically new magnetic recording system designs, the utility of and need for a versatile and reliable guide among the many options and directions is apparent. This proposal builds on research previously funded by NSF through the NCR and ECS programs, and funded by NSIC, ARPA, Department of Commerce (ATP), and industry to put together a systems-oriented research program in magnetic information systems.Central to much of this work is the measurement, modeling, and analysis of waveforms resulting from the magnetic microstructure of the medium. This microstructure effectively processes the written signal, and the synthesis of new systems should be guided in part by this processing. The medium magnetization is peak-limited, nonlinear, hysteretic, and granular. The very ability to retain information recorded also ultimately limits a medium's potential performance. The dependence of system performance on varying fundamental medium parameters such as exchange coupling, anisotropy energy, granularity, and film thickness will be pursued. Using more powerful analytical tools to model these medium effects, the performance of systems designed to compensate for repeatable medium noise effects will be quantified. Signal degradation due to applied field, time, and temperature, is quantified using unique experiments developed using the magnetic force microscope arid recording testers at Washington University.A physically based simulation of magnetic recording systems, including media, read heads, and write heads has been developed. This computational model forms a bridge between theoretical and experimental research. Many of its features have been verified by comparison of its simulations with VSM and spin stand measurements, including some fairly detailed recording procedures. The model has now reached a stage of development where it is envisaged as a tool for system synthesis. As the tool is refined, it must be kept modular so that new head designs, new media, and new signal processing blocks may be incorporated easily, and their interactions evaluated. It must retain an ability to simulate single realizations of media and corresponding waveforms, and to run Monte-Carlo simulations to predict system level performance for a variety of performance metrics. This model will be expanded: to predict signal dependence on applied fields, time, and temperature; to include current communications signal processing techniques used and proposed for magnetic recording channels; to include perpendicular and patterned media models; and to incorporate models for probes such as the magnetic force microscope.***