RIA: Development of a Low-Cost, Near-Optimal Detector for Magnetic Data Storage Systems

在数据存储系统中，采用先进的信号处理技术作为实现高存储密度的一种经济有效的手段变得越来越重要。首席研究员对磁数据存储通道的实用序列检测算法的研究导致了一种计算效率高的检测算法的成功开发，该算法在磁记录中重要的调制码约束下实现了近乎最佳的性能。在这个项目中，首席研究员解决了与所提出的检测算法应用于实际记录系统有关的几个重要问题。这些问题涉及：1)与所提出的检测方案相关的固有误差传播问题；2)代码和检测器对实际磁记录通道中经常遇到的各种干扰的灵敏度；3)硬件实现的不同选项；4)自适应制定算法，并开发了时变信道环境所需的高级时序恢复方案。该项目的目标是仔细检查这些问题，在实时变化的信道环境下确定有利的硬件设计选项，并在潜在问题存在时提供解决方案。与传统方法相比，所提出的方法的潜在好处和局限性将被彻底理解。除了提供与硬件实现直接相关的信息外，所提出的研究结果还有助于我们理解不同类型信道干扰下的代码和检测器行为。这将为追求设计不受选定类型信道干扰影响的代码和检测器的挑战性问题提供基础。

The use of advanced signal processing techniques is becoming increasingly important as a cost-effective means to achieve high storage density in data storage systems. The Principal Investigator's research on pragmatic sequence detection algorithms for magnetic data storage channels has resulted in the successful development of a computationally efficient detection algorithm, which achieves a near-optimum performance under modulation code constraints that are important in magnetic recording. In this project, the Principal Investigator addresses several important issues related to the application of the proposed detection algorithm to practical recording systems. These are the issues related to: 1) the inherent error propagation problem associated with the proposed detection scheme, 2) code and detector sensitivity against the types of disturbances often encountered in real magnetic recording channels, 3) different options available for hardware implementation, and 4) adaptive formulation of the algorithm and the development of an advanced timing recovery scheme necessary for time-varying channel environments. The goal of this project is to carefully examine these issues, determine favorable hardware design options under real time-varying channel environments and provide solutions whenever potential problems exist. Potential benefits and limitations of the proposed method in comparison with conventional approaches will be thoroughly understood. In addition to providing information that is directly related to hardware implementation, the results of the proposed research help us understand code and detector behaviors under different types of channel disturbances. This will provide a basis for pursuing the challenging problem of designing codes and detectors which are immune to a selected type of channel disturbance.