|
2 Day Course
This is the ideal course for anyone who needs to quickly come up to speed on today's data storage read channels. It focuses on the detection methods and practical considerations that lead from simple peak detection ideas to today's Noise Predictive Maximum Likelihood (NPML), or Noise Predictive Viterbi (NPV), data detectors employing Parity-Assisted Post-Processing.
You will understand the difference between PR (partial response) and ML (maximum likelihood sequence detection) and how they work together to form the industry standard of data detection. Plus you will gain an intuitive understanding of correlated noise, its causes, and how it affects NPML/NPV channels. You will also understand key practical considerations such as data-dependencies, non-linearities, detector target optimization, recording codes, gain and timing recovery and filter adaptation, as well as how the system works and its tradeoffs. Additionally, many attendees from the tape and optical disc industries have found the course material and insights provided useful and applicable to their channels, as well. We also briefly introduce emerging topics such as turbo codes, LDPC codes, iterative detection, turbo equalization and the problem of low SNR timing recovery.
The course is designed for those new to recording channels, while providing industry veterans with a clear, concise review of key topics and the insight and context needed to understand and assess new developments. Suitable industry experience or previous disk drive technology training will satisfy the prerequisites.
Recording Process Overview
Read and Write Process
NRZ and NRZI Data Sequences
Pulse Models
Intersymbol Interference (ISI)
User Density
Read Channel Basics
The Typical Drive Layout
Data Rate
Error Mechanisms
White Gaussian Noise
For Comparison:
Sampled Peak Detection
Calculating BER Performance
Maximum Likelihood
Sequence Detection
The Viterbi Algorithm Revealed!
Path Metric Calculation
Simplifications
Signal Non-Linearities
What is Linearity?
Non-Linear Transition Shift (NLTS)
Media Noise (Transition Jitter)
Partial Erasure
Pulse Asymmetry
Thermal Asperities
Signal Processing Basics
Sampling Bandlimited Signals
Aliasing
Transforms: D, Fourier
Sinc Function
Filtering and Modeling
Binary Arithmetic:
Pre- and Post-Coding
Traditional Partial Response
Targets
Developing Intuition
Class 4 Partial Response
Traditional PR4, EPR4, EEPR4
Targets
'Natural Density'
The 'Ideal' Target? |
Error Events in Sequence Detectors
(Minimum) Euclidean Distance
Calculating BER Performance
What Causes an Error?
Error Event Notation
Gain and Timing Recovery: AGC & PLL
The Effects of ISI
Preliminary Target Determination
Timing Functions
Illustrations of Timing Function Behavior
Introduction to
Interpolated Timing Recovery (ITR)
Adaptive Equalization
Target Spectra
Designing the Equalizer
Continuous-Time Filters
LMS Adaptation of FIR
Interaction of Feedback Control Loops
PRML, GPR, NPML, NPV
Traditional PRML State Diagrams
and Trellises
Methods for Comparing Channels
Introduction to Noise Prediction
and Non-Standard Targets
Noise Whitening Filters
Non-Linear Transition Jitter Noise Model
SNR Definition in Media Noise
Matched Filtering and Its Effects
on Noise Correlation
and Detector Targets
Coding
Run-Length Limited (RLL) Code Basics
d=0 Codes for PRML
Introduction to Distance Enhancing Codes
Quantifying Error Events
Maximum Transition Run (MTR) Codes
Rate-1 Codes?
Permuted Architectures
Conclusions and Future Trends
The Big Picture
Parity-Assisted Post-Processors
The Next dB
Introduction to LDPC, Turbo Codes,
Iterative Detection, Turbo Equalization,
and Low SNR Timing Recovery
Perpendicular Recording
Effects of Emerging Market Segments
New Challenges
|
Click on a course date to register for that session
Click on the location to view seminar site information
Home | Info | Courses | Schedule | Instructors | RegisterFor questions or comments about this web site please contact the webmaster@KnowledgeTek.com |
.jpg)
