# ADVANCED SIGNAL PROCESSING - 2018/9

Module code: EEEM007

Module provider

Electrical and Electronic Engineering

Module Leader

PLUMBLEY MD Prof (Elec Elec En)

Number of Credits

15

ECTS Credits

7.5

Framework

FHEQ Level 7

JACs code

H690

Module cap (Maximum number of students)

70

Module Availability

Semester 2

Overall student workload

Independent Study Hours: 105

Lecture Hours: 33

Laboratory Hours: 12

Assessment pattern

Assessment type | Unit of assessment | Weighting |
---|---|---|

Coursework | LABORATORY & ASSIGNMENT | 30 |

Examination | EXAMINATION - 2HRS | 70 |

Alternative Assessment

Not applicable: students failing a unit of assessment resit the assessment in its original format.

Prerequisites / Co-requisites

None.

Module overview

Expected prior learning: Knowledge of the basic principles of matrices and vectors; probability and statistics; and digital signal processing (Fourier Transforms, LTI Systems, Z-Transforms). MEng students might have acquired this by study of modules EEE2035 Engineering Mathematics III and EEE3008 Fundamentals of Digital Signal Processing. MSc students might have acquired this by study of a module in engineering mathematics, and module EEE3008 Fundamentals of Digital Signal Processing or similar.

Module purpose: Advanced signal processing, which includes adaptive filtering, signal detection, matching and recognition, is a key expertise required for designing and building high-tech. electronic systems such as robots, automatic speech recognition systems, driver warning systems, biometrics technology, etc. The module will introduce students to advanced techniques of signal processing and interpretation.

Module aims

Equip students with advanced analytical tools for solving the statistical and adaptive signal processing problems encountered in communications, telematics, and related engineering areas; and

To introduce students to statistical and adaptive techniques for the detection, filtering and matching of signals in noise;

Make students aware of the industrial relevance of these techniques.

Learning outcomes

Attributes Developed | ||
---|---|---|

1 | Explain the concepts and theory of statistical and adaptive signal detection, filtering and matching | K |

2 | Demonstrate the ability to apply mathematical models of signal processing to solve problems and predict effects??cx00 | C |

3 | Describe the relevance of the presented material to applications in machine perception and discuss its engineering significance | P |

4 | Design pattern-recognition systems | KCP |

5 | Demonstrate technical expertise required for performance characterisation of machine perception systems | KCPT |

Attributes Developed

**C** - Cognitive/analytical

**K** - Subject knowledge

**T** - Transferable skills

**P** - Professional/Practical skills

Module content

Indicative content includes the following:

Lecture Component Statistical Pattern Recognition

Hours: 22 Lecture/Tutorial hours

Review of linear algebra; review of probability theory.

Elements of Statistical Decision Theory - Model of pattern recognition system. Decision theoretic approach to pattern classification. Bayes decision rule for minimum loss and minimum error rate. Discriminant Functions.

Supervised learning. Learning algorithms. Classification Error Rate Estimation.

Nearest Neighbour (NN) Technique - 1-NN, k-NN pattern classifiers. Error bounds. Editing techniques. Probability Density Function Estimation - Parzen estimator, k-NN estimator.

Unsupervised Learning and Cluster Analysis - Concepts of a cluster, k-means algorithm, hierarchical clustering.

Feature Selection - Concepts and criteria of feature selection. Algorithms for selecting optimal and sub-optimal sets of features. Recursive calculation of parametric separability measures.

Feature Extraction - Probabilistic distance measures in feature extraction. Properties of the Karhunen-Loeve expansion, feature extraction techniques based on the Karhunen-Loeve expansion. Discriminant analysis.

Classifier Fusion - Fusion System architecture. Fusion rules and their properties

Lecture Component: Adaptive Digital Filtering

Hours: 11 Lecture/Tutorial hours

Introduction – Review of Fourier Transform, Z-Transform, LTI system analysis.

Approaches to adaptive filters. State space model. Cost functions.

Correlation matrix, autoregressive and moving - average models.

Spectral analysis.

Linear Prediction.

Mean Square Estimation - Conditional expectation and orthogonality. Wiener filtering.

FIR Adaptive Filters.

Methods of Teaching / Learning

The learning and teaching strategy is designed to achieve the specified learning outcomes by teaching the module syllabus in lectures, and supporting the assimilation and understanding of the taught material via tutorial classes. The practical design and technical skills related to the subject are acquired through coursework involving an assignment on pattern recognition system design; the performance of the designed pattern recognition system is characterised via laboratory experiments.

Learning and teaching methods include:

Lectures: 10 weeks, 2-3 hours per week

Tutorials: 6-8 weeks, 1 hour per week

Revision classes: 2-3 hours

Assignment: Pattern recognition system design (linked to Pattern Recognition Lab)

Labs: Pattern Recognition Experiment (Preparation via assignment)

Assessment Strategy

The assessment strategy for this module is designed to provide students with the opportunity to demonstrate the learning outcomes. The written examination will assess the knowledge and assimilation of the terminology, concepts and theory of statistical and adaptive signal processing, as well as the ability to analyse problems and apply mathematical models of signal processing to solve and predict effects. The Assignment will assess the ability to design pattern recognition systems. The laboratory experiment will evaluate the acquired technical skills and expertise required for performance characterisation of pattern recognition systems.

Thus, the summative assessment for this module consists of the following.

2-hour, closed-book written examination;

Pattern recognition system design: An assignment involving the design of a pattern classification system;

Pattern Recognition Experiment: A laboratory experiment concerned with pattern recognition system evaluation

Formative assessment and feedback

The main mechanism for formative assessment will be a set of tutorial problem sheets which the students will be expected to solve prior to the timetabled tutorial classes. The tutorial classes will provide the main forum for formative feedback. In addition, formative assessment and feedback may occur:

During lectures, by question and answer sessions;

During tutorials/tutorial classes;

During supervised laboratory sessions;

During meetings with lecturers and tutor.

Reading list

Reading list for ADVANCED SIGNAL PROCESSING : http://aspire.surrey.ac.uk/modules/eeem007

Programmes this module appears in

Programme | Semester | Classification | Qualifying conditions |
---|---|---|---|

Computer Vision, Robotics and Machine Learning (EuroMasters) MSc | 2 | Compulsory | A weighted aggregate mark of 50% is required to pass the module |

Computer Vision, Robotics and Machine Learning (EuroMasters) MSc | 2 | Compulsory | Each unit of assessment must be passed at 50% to pass the module |

Communication Systems MEng | 2 | Optional | A weighted aggregate mark of 50% is required to pass the module |

Computer and Internet Engineering MEng | 2 | Optional | A weighted aggregate mark of 50% is required to pass the module |

Electronic Engineering MEng | 2 | Optional | A weighted aggregate mark of 50% is required to pass the module |

Electronic Engineering with Computer Systems MEng | 2 | Optional | A weighted aggregate mark of 50% is required to pass the module |

Computer Vision, Robotics and Machine Learning MSc | 2 | Optional | A weighted aggregate mark of 50% is required to pass the module |

Mobile Media Communications MSc | 2 | Optional | A weighted aggregate mark of 50% is required to pass the module |

Electronic Engineering (EuroMasters) MSc | 2 | Optional | A weighted aggregate mark of 50% is required to pass the module |

Electronic Engineering MSc | 2 | Optional | A weighted aggregate mark of 50% is required to pass the module |

Communications Networks and Software MSc | 2 | Optional | A weighted aggregate mark of 50% is required to pass the module |

Mobile and Satellite Communications MSc | 2 | Optional | A weighted aggregate mark of 50% is required to pass the module |

Mobile Communications Systems MSc | 2 | Optional | A weighted aggregate mark of 50% is required to pass the module |

Communications, Networks and Software (EuroMasters) MSc | 2 | Optional | A weighted aggregate mark of 50% is required to pass the module |

Mobile and Satellite Communications (EuroMasters) MSc | 2 | Optional | A weighted aggregate mark of 50% is required to pass the module |

Mobile Communications Systems (EuroMasters) MSc | 2 | Optional | A weighted aggregate mark of 50% is required to pass the module |

Mobile Media Communications (EuroMasters) MSc | 2 | Optional | A weighted aggregate mark of 50% is required to pass the module |

Electronic Engineering with Communications MEng | 2 | Optional | A weighted aggregate mark of 50% is required to pass the module |

Electronic Engineering with Audio-Visual Systems MEng | 2 | Compulsory | A weighted aggregate mark of 50% is required to pass the module |

Please note that the information detailed within this record is accurate at the time of publishing and may be subject to change. This record contains information for the most up to date version of the programme / module for the 2018/9 academic year.