Optional Learning Units

Under construction!

The following courses are available for 2016/2017. Students may choose any combination of three courses.

 
 
 
 
Semester 2
 
 
Option 1 
 
Option 2
 
Option 3
Data Processing and Modelling *
Mobile Communication Systems *
Optical Communications *
RF Circuits and Subsystems *
Wireless Networks and Protocols *

 

Data Processing and Modelling

Lecturers: Armando J. Pinho, Pétia Georgieva, Telmo Reis Cunha

Aims: This course is to gather knowledge in advanced areas of Data Driven Modelling and it consists of the following three teaching modules: A) Data compression and data complexity, B) Machine Learning, and C) Nonlinear System Identification.

A) Data compression and data complexity

Module Summary: This module aims at providing a brief overview of the information theory behind discrete source coding and the corresponding data compression algorithms. Data compression is among the most important research areas in modern computing. In fact, representing information efficiently is a key requirement in virtually every system, including those handling text, audio, image or video, but also other types of information, such as DNA sequences. The main objective of data compression is to represent messages with less bits than those originally used. However, lossless data compression also plays an important role in measuring the complexity of data. This is a less known use of data compression algorithms, but that is gaining increasing interest, for example for assessing the complexity of DNA sequences, for measuring the similarity between images, or for discovering the authorship of a certain written text. In this module, we will address and explore this important connection.

Training outcomes: Knowledge of important aspects of data compression and data complexity and their relations with information theory and Kolmogorov complexity.

Program:

1. Basic concepts of information theory

Entropy
Joint entropy and conditional entropy
Relative entropy
Mutual information

2. Variable-length codes

The Kraft inequality
Optimal codes
Huffman codes

3. Universal source codes

Arithmetic coding
Lempel-Ziv coding

4. Data compression and data complexity

Basic notions on the theory of computability
Kolmogorov complexity
The normalized compression distance
Other compression-based measures Applications

B) Machine Learning

Module Summary: Machine Learning (ML) is a subfield of artificial intelligence that is concerned with the design, analysis, implementation, and applications of programs that learn from examples or experience. Learning from data is commercially and scientifically important. ML consists of methods and respective software that extract automatically interesting knowledge (patterns, models, relationships) in large databases of sometimes chaotic and redundant information. ML is a data-based knowledge-discovering process that has the potential not only to analyze events in retrospect but also to predict future events or important alterations. It is a top field in computer science with growing job options.

Training outcomes: This module will provide an introduction in the field of Machine Learning. Upon successful completion of the module, students will have a general understanding of ML algorithms and their use in big data mining. Students will have designed and implemented several ML algorithms. Students will also be able to identify, formulate and solve ML problems that arise in practical applications. Students will have knowledge of the strengths and weaknesses of different ML algorithms (relative to the characteristics of the application domain) and be able to adapt or combine some of the key elements of existing ML algorithms to design new algorithms as needed.

Program:

I. Fundamentals of Machine Learning

1. Supervised learning

Linear (univariate/ multivariate) regression
Logistic regression. Regularization
Artificial Neural Networks (ANN)
Support Vector Machines (SVM)

2. Unsupervised learning

K-means clustering
Data dimensionality reduction
PCA (Principal components analysis)

3. Reinforcement learning

II. Machine Learning applications

1. Handwritten digits recognition (from images)
2. Spam e-mail recognition
3. Biomedical data classification


C) Nonlinear System Identification

Module Summary: This module discusses state-of-the-art tools used to construct mathematical representations of the input- output behavior of a nonlinear dynamic system (behavioral models), a process commonly known as System Identification. Achieving mathematical representations of system behavior is fundamental for many different system analyses such as simulation, control, design, sensitivity analysis, and others. This module is divided in two main parts. The first one, after introducing the basic concepts of System Identification, is focused on the main model topologies and structures that are used for modeling linear and nonlinear dynamic systems, where the latter receives the strongest emphasis. Universal approximator models – polynomial filters and artificial neural networks – are thoroughly analyzed. Beyond the presentation and support of each model, issues such as model complexity, implementation, predictive ability, accuracy, robustness and simulation efficiency are thoroughly addressed. The second part is dedicated to model parameter extraction methodologies, with the presentation of the most important methods used to calculate the parameters of linear time-invariant, nonlinear static and nonlinear dynamic models. Parametric and non-parametric system identification approaches are explored through the presentation of time- and frequency-domain methods. Nonlinear optimization, correlation analysis and least-squares estimation methods are presented in detail in this part. Practical applications of system identification techniques to a set of test-cases (of distinct nature) are presented along with the theory throughout the module, illustrating the concepts lectured in the module classes.

Training outcomes: At the end of this module the students should:

  • have a clear idea of the importance and necessity, in scientific activities, to model systems;
  • have knowledge of the main modeling structures and topologies that are available to model the input-output behavior of a system;
  • be able to implement such model topologies and algorithms to simulate them, using mathematical and signal processing software tools (such as MATLAB);
  • to understand and implement the strategies used to estimate the values of the model parameters, recognizing the key issues that are determinant in this step (which will have impact on the model accuracy);
  • be able to implement in a software tool the whole process for modeling a nonlinear dynamic system (from model selection, implementation, extraction, simulation, accuracy estimation and robustness evaluation).

Program:

1. Basics on System Identification Theory

a. Introduction – An overview of mathematical representation
b. Behavioral (input-output) versus Physical (phenomenological) Modeling

2. Model Structures for Time-Invariant Systems

a. General Nonlinear Dynamic TI System – Universal and Non-Universal Approximations
b. Parametric and Non-Parametric System Models in Time and Frequency-Domains
c. Recursive and Non-Recursive Models
d. Universal Approximators: Polynomial (Volterra) Filters and Time-Delay Artificial Neural Networks
e. Non-Universal Approximators: The Two- and Three-Box-Model Cascades, Two-Box Feedback Models, and others

3. Model Extraction and Validation

a. Behavioral Modeling Extraction and Excitation Design
b. Parameter Extraction through Least-Squares
c. Correlation Methods
d. Nonlinear Optimization – The Error Back-propagation Algorithm
e. Maximum Likelihood Method

 

Mobile Communication Systems

Lecturers: Atílio Gameiro (coordinator), António Navarro

Learning outcomes:

  • ability to identify and understand the main requirements, issues, limitations, parameters and components used in the design of point to point and multi-user radio links;
  • ability to understand the rationale for the solutions adopted in existing and emerging systems;
  • increased sensibility to cutting-edge topics in future wireless / mobile systems, translated in the capacity to participate in the development and proposal of new techniques aiming to answer the goals foreseen for future systems;
  • ability to extract information from scientific papers in the area. • Technical writing and presentation skills.

Syllabus:

Component One

I. Introduction and basic principles 

1. Introduction

Mobile and wireless devices
Components of a radio system
Electromagnetic spectrum

2. Wireless networks

The different types of wireless networks (cellular, WLAN, broadcasting,...)
Current wireless Systems
Spectrum allocation for current wireless systems

II. Wireless Channel 

Propagation effects that affect the communication
Parameters used to characterize wireless channels
Path loss, shadowing, fast fading
Statistical characterization of multipath
Review of digital modulation types
Performance of modulation schemes in fading channels
Capacity of the SISO, MISO and MIMO wireless channels

III. Principles of OFDM and DVB-T/T2 Systems 

OFDM main advantages and disadvantages
IDFT and DFT
Guard interval
Mode selection
Single frequency
PAPR reduction techniques
MISO

IV. Advanced channel coding for modern wireless systems 

Shannon capacity theorem
Convolution codes and concatenated codes
Puncturing
Principles of turbo coding
RSC codes
LDPC

V. Multiple access schemes

1. Classification and usage of multiple access schemes

TDMA, FDMA, CDMA, hybrid schemes
Usage in the diferente generation of cellulat systemas

2. OFDM based multiple access schemes

Multicarrier signal design
Review of the OFDM concept
Performance of OFDM over the fading channel
Coding for OFDM
OFDMA concept and issues
Single carrier frequency division ultiplexing (SC-FDM)
New waveforms proposed for 5G

VI. Link and multilink design concepts for fading channels

1. Concept and forms of diversity 

Realization of independent fading paths
Basic diversity forms
Introduction to multiple antenna and spatial diversity

2. Multiple antenna systems

Benefits offered by multiple antenas
Space diversity, capacity, beamforming
Diversity schemes with multiple antenas
Transmit / receive diversity
Space-time-frequency codes
Capacity enhancement with multiple antenas
Multiplexing vs beamforming
Capacity of MIMO systems
MIMO design for multiplexing
Trade offs capacity vs diversity 

3. Multiuser diversity

Concept
Multi-user diversity through relaying

Component Two 

This component will tackle topics that represent trends and emerging techniques in the field of wireless / mobile communication systems. The topics for this component will be proposed at the beginning of the course will be assigned as homework for student research and presentation. Eventually some topics can be presented through guest lectures. Examples of possible topics are:

Distributed antenna systems
Network coding
Integration of radar and communications
Novel waveforms for 5G
Novel modulations

 

Optical Communications

Lecturers: Mário Lima (Coordinator), António Teixeira, Luís Pessoa, Henrique Salgado

AimsThe course aims to provide the students with the fundamentals related to optical communication systems and networks, presenting nowadays scenarios (core and access), and foreseeing next generation optical networks (NGN).

It discusses several issues covering in a first part the principles of optoelectronics and fiber optics operation, followed by optical networks aspects, namely related to access passive optical networks (NGA-PONs): standards, design and installation.

The students will be able to receive complementary laboratory formation, by performing some experiments related to the course topics.

Program:

1. Optical communication systems fundamentals

1.1. Optical fibers

1.1.1. Geometric theory of light propagation in multimode fibers
1.1.2. Mode theory of light propagation in single-mode fibers
1.1.3. Attenuation mechanisms
1.1.4. Dispersion (intermodal, chromatic)
1.1.5. Fiber types (SMF, DSF, DCF)
1.1.6. Scattering Mechanisms
1.1.7. Fiber limitations

1.2. Optical Sources

1.2.1. Operation basics
1.2.2. LEDs and laser diodes
1.2.3. Noise and chirp
1.2.4. Fiber limitations (bit rate x distance) 1.2.5. Intensity modulation (direct and external)

1.3. Optical amplifiers

1.3.1. Different types of amplifiers (EDFA, Raman, SOA)
1.3.2. EDFA operation (gain and noise)
1.3.3. EDFA in WDM systems

1.4. Photodiodes and receivers

1.4.1. PIN and APD
1.4.2. Noise sources
1.4.3. Pre-amplified receivers
1.4.4. Direct and coherent detection

1.5. Nonlinear effects in fiber

1.5.1. SPM, XPM
1.5.2. Pulse propagation

1.6. Advanced modulation formats

1.6.1. Mach-Zehnder modulator
1.6.2. Formats (transmitter, receiver, spectral characteristics)

2. Next generation optical networks

2.1. Nowadays scenarios (competing technologies)

2.1.1. Metro networks
2.1.2. Local networks

2.2. Core/metro networks

2.2.1. Components involved
2.2.2. Design

2.3. Access networks (NGA)

2.3.1. Components involved
2.3.2. Topologies FTTx
2.3.3. Access T echnologies
2.3.4. Standards (xPON)
2.3.5. Design

2.4. NGN in Portugal/world


RF Circuits and Subsystems

Lecturers: José Machado da Silva, Nuno Borges Carvalho, Paulo Pereira Monteiro

AimsThis course is focused on RF electronic circuits and subsystems and is intended to complement the basic undergraduate knowledge on telecommunications electronics of PhD students, as RF systems are the basis of all wireless systems. These currently include all types of mobile communication systems, Bluetooth, Zig-bee and WiFi, and will be the basis for future solutions on 5G communications, especially Internet of Things, white space technologies and M2M.

Pre-requisites: It is expected that students have already undergraduate knowledge of electromagnetism, electromagnetic wave propagation in guided media and electronic circuits.

Summary of the expected learning outcomes at the end of the UC:

1. Understanding the basics of RF circuits and their operation.
2. Understand the different measurement units that are used to measure RF signals.
3. Allow students to understand the specificities of wireless measurements.
4. Understand how test and measurement (T&M) equipment’s are internally build.
5. Test and measure typical wireless components.
6. Hybrid and monolithic microwave integrated circuit design.
7. Design of electronic RF circuits.
8. Knowledge of radio-frequency communication systems with emphasis on implementations and practical aspects related to transceivers.

Program:

1. Measurement of wireless transceivers 

1.1. Introduction
1.2. Linear two-port networks
1.3. Linear FOMs
1.4. Nonlinear two-port networks
1.5. Nonlinear FOMs
1.6. System level FOMs
1.7. Filters
1.8. Amplifiers
1.9. Mixers
1.10. Oscillators
1.11. Frequency multiplier FOMs
1.12. Digital converters

2. Instrumentation for wireless systems

2.1. Introduction
2.2. Power meters
2.3. Spectrum analyzer
2.4. Vector signal analyzer
2.5. Real time signal analyzer
2.6. Vector network analyzer
2.7. Nonlinear vector network analyzer
2.8. Oscilloscopes
2.9. Logic analyzer
2.10. Noise figure measurement

3. Signal excitation

3.1. Introduction
3.2. One tone excitation
3.3. Two tone excitation
3.4. Digitally modulated signals
3.5. Chirp signals
3.6. Comb generators
3.7. Pulse generators

4. Test benches for wireless systems characterization and modelling

4.1. Introduction
4.2. Test benches for characterization
4.3. Test benches for behaviour modelling


Wireless Networks and Protocols

Lecturers: Adriano Moreira (coordinator), Manuel Ricardo, Rui Aguiar

Aims: Wireless Networks and Protocols (WNP) is a course for students aimed at specializing in the mobile communications theme of MAP-Tele. The WNP course has two main objectives:
1. to provide the students with the competences required to understand current wireless networks and their main functions;
2. to provide students with the competences required to create future wireless networks and/or its associated functions.
In order to meet these objectives a set of scientific topics were identified: a) wireless networking, b) mobility, c) authentication, d) Quality of Service (QoS), and e) network support for services.

Students should be able to:

1. Describe the evolution path of the wireless and mobile communications systems;
2. Enumerate the most relevant wireless communications technologies and identify the corresponding standards and major players;
3. Explain how the current wireless network systems cohabit;
4. Describe the major technical capabilities and limitations of the current wireless communications systems;
5. Identify current trends in telecommunications systems integration, from networks to service support;
6. Describe the emerging paradigms in communications networks integration;
7. Explain the importance of authentication and access control mechanisms in integrated wireless networks;
8. Describe the most relevant models for the support of quality-of-service in wireless networks, and identify the challenges for their implementation;
9. Describe the concept of service oriented architectures and provide examples of solutions.

Syllabus:

1. Introduction to Wireless Networks and Protocols: a) Overview; b)History; c)Standards and market issues; d)Evolution and trends.

2. Fundamentals of wireless communications: a) Transmission; b) Wireless data links and medium access control; c; Networking; d) Mobility concepts; e) Research issues.

3. Telecommunications systems: a) GSM; b) GPRS; c) UMTS; d) LTE; e) TETRA; f) Broadcast and satellite.

4. IEEE wireless data networks: a) WPAN; b) WLAN; c) WMAN.

5. Convergence and interoperability: a) Evolution of 3GPP networks; b) Wireless mesh networks; c) Research issues.

6. Quality of service: a) Characterization and models; b) Case studies: 3GPP-QoS, IEEE-QoS, IP-QoS; c) Research issues.

7. Support for services and applications: a) Web services components; b) Services and applications platforms; c) Research issues.

8. Authentication and access control: a) Fundamentals of Authentication and Access Control; b) Characterization and models; c) Case studies: 3GPP, 802.1x, 3GPP; d) Research issues.


Comments