Development of Statistical Signal Processing Algorithms
This 3-day course covers the fundamental approaches to developing statistical signal processing algorithms to meet system design specifications. Additionally, the algorithms that are currently used in practice and have stood the “test of time” are highlighted. The methodology to design, evaluate, and test new algorithms is presented in a simple step-by-step manner. In doing so, the computer language MATLAB is utilized. All algorithms and methods discussed have been implemented in MATLAB and will be provided to the attendee. The course is designed for engineers, scientists, and other persons who wish to implement and/or design statistical signal processing algorithms without having to scour the current literature for possible solutions. The presentations will emphasize actual working algorithms and will deemphasize the mathematics behind them so that the course will be accessible to those who may not be familiar with the theoretical foundations. Knowledge of the instructor’s previous books, “Modern Spectral Estimation”, “Fundamentals of Statistical Signal Processing: Estimation” and “Fundamentals of Statistical Signal Processing: Detection” is not required. Attendees are encouraged to bring their laptops so that they are able to exercise the programs along with the instructor. All MATLAB source code will be provided for course and future use. Each participant will receive the recently released book “Fundamentals of Statistical Signal Processing, Vol. III, Practical Algorithm Development”, by Steven Kay, Prentice-Hall, 2013. The book contains a CD with the MATLAB programs.
What you will learn:
- Step by step approach to the design of algorithms
- Comparing and choosing signal and noise models
- Performance evaluation, metrics, tradeoffs, testing, and documentation
- Optimal approaches using the ‘big theorems”
- Algorithms for estimation, detection, and spectral estimation
- Lessons learned and “rules of thumb” for each topic
- Complete case studies
- Methodology for algorithm design: flow charts, example of algorithm design
- Mathematical modeling of signals: linear vs. nonlinear, deterministic signals, random signals, unknown parameters
- Mathematical modeling of noise: white Gaussian noise, colored Gaussian noise, general Gaussian noise, IID nonGaussian noise
- Signal model selection: flow charts, random processes, transients, periodic models, model order estimation
- Noise model selection: flow charts, estimation of probability density functions, spectrum, moments, covariance matrix, model order estimation, confidence intervals
- Performance, evaluation and testing: metrics, Monte Carlo evaluations, bias versus variance, mean square error, probability of error, receiver operating characteristics, software development, documentation
- Optimal approach using the big theorems: Neyman-Pearson, likelihood ratio, maximum likelihood, maximum a posterior, minimum MSE, linear models
- Specific algorithms for estimation, detection, and spectral estimation: parameter estimation, signal extraction, adaptive filtering, sinusoidal estimation, matched filters, estimator-correlator, spectral estimation via Fourier and high resolution methods
- Complex data extensions: complex demodulation, complex random variables and random processes, extensions of all algorithms to complex data
- Case studies: Radar Doppler center frequency estimation, magnetic signal detection, and heart rate monitoring
This course is not on the current schedule of open enrollment courses. If you are interested in attending this or another course as open enrollment, please contact us at (410) 956-8805 or at email@example.com and indicate the course name and number of students who wish to participate. ATI typically schedules open enrollment courses with a lead time of 3-5 months. Group courses can be presented at your facility at any time. For on-site pricing, request an on-site quote. You may also call us at (410) 956-8805 or email us at firstname.lastname@example.org.
Steven M. Kay is one of the world’s leading experts in statistical signal processing. Currently Professor of Electrical Engineering at the University of Rhode Island, Kingston, he has consulted for numerous industrial concerns, the Air Force, Army, and Navy and has taught short courses to scientists and engineers at NASA and the CIA. Dr. Kay is a Fellow of the IEEE. He has received the Education Award for “outstanding contributions in education and in writing scholarly books and texts…” from the IEEE Signal Processing society and has been listed as among the 250 most cited researchers in the world in engineering.
Contact this instructor (please mention course name in the subject line)