Data Processing for Expeditionary Warfare (USA)

Course structure

What you will learn

On successful completion of this module a student should be able to:

Knowledge and Understanding

• Describe and analyze the principles, capabilities and limitations of a number of different sensor processing methods, algorithms and schemes
• Critically assess performance of neural network models
• Evaluate computational models for dealing with uncertainty; discern the utility of Bayesian and Fuzzy models

Skills and Other Attributes

• Compare and contrast different estimation methods and tools when applied to a specific problem and devise a distributed or centralized processing scheme based on this comparison
• Design, develop and implement models in MATLAB

Core content

• Adaptive Signal Processing: Adaptive FIR and spatial filters, error surface, Newton’s method, gradient search method, LMS algorithm, practical examples
• Multi-layer perceptrons: Architecture, Back-propagation algorithm, performance of the algorithm, Unsupervised learning, Hebbian learning, Kohonen maps, Neural network design in Matlab
• Overview of deep learning algorithms
• onar signal processing: Beamforming, passive analysis, active processing
• Fuzzy logic: fundamentals, fuzzy associative matrix, fuzzy inference
• Adaptive linear elements: Tapped delay lines, Noise cancellation, Time series prediction,
• Sensor fusion: distributed sensor systems, sensor coverage, distributed vs centralised systems
• Sensor processing: Genetic algorithms, evolutionary algorithms, artificial neural nets
• MATLAB: practical sessions, demonstrating the above concepts in Matlab