Master of Science in Embedded Systems Design

Embedded Systems Design

The Master of Science in Embedded Systems Design is promoted by ALaRI (Advanced Learning and Research Institute), in association with Faculty of Informatics of the University of Lugano. It is offered in collaboration with ETH Zurich and Politecnico di Milano. Goals and Contents Embedded systems offer challenging opportunities to young designers, both by interweaving technologies such as microelectronics, computer science and telecommunications, and by focusing on the most advanced applications. Aware of the real need for a cross-disciplinary approach to education, ALaRI equips talented students with a unique body of knowledge about the multi-interdisciplinary area of embedded systems. Training includes interpersonal skills indispensable in today’s industry, such as team work, complex project management and market sensitivity. ALaRI research activities focus on topics of great scientific interest and industrial applicability, based on real-life design issues. The main research trends concern: Security, Pervasive Computing, System-level Design, System-on-chip. The degree programme stretches over two years with four

164

semesters of study organized in mandatory and elective courses, including the development of the final Master’s project. Students can personalise their study curricula according to their individual interests, expectations and ambitions. Design laboratories follow an interdisciplinary approach in which students learn the discipline of design, applying the competences acquired in the different courses in an integrated way. A research project is an integral part of this study programme, leading to the final Master’s thesis. During the first two semesters, the study programme is identical for all students; subsequently, students can choose between two tracks: – Design and Research - towards a career in either the academic or the industrial environment – Business Projects – towards a career involving also management, economics and marketing of embedded systems Both tracks qualify graduates to seek admission to PhD studies. The ALaRI teaching programme is characterized by two distinct features: – cooperation between university and industry – interdisciplinarity through design laboratories and projects Professors from renowned European and US universities and research centres assure top-quality teaching. Experts from industry contribute with their know-how as tutors or advisors in the research projects.

165

Master of Science in Embedded Systems Design

Sem. 3+4

Sem. 1+2

Study Programme* First year Common Courses (60 ECTS) Introduction Courses Technical Courses** Second year Design and Research (60 ECTS) Technical Courses** Master Project Design Lab 2

Total ECTS

0 50

Design Lab 1 Master Project***

40 15 5

Second year Business Projects (60 ECTS) Technical Courses** Design Lab 2 Business Courses Business Plan on Master Project Master Project***

5 5

28 3 12 3 14 120

After the second semester, students can choose between two tracks: Design and Research or Business Projects.

* Slight changes in the study programme may occur. ** Technical courses consist of fundamental and elective courses. *** The Master Project starts during the second semester, and continues during the third and fourth ones, leading to the final Master Thesis.

Master of Science in Embedded Systems Design

166

Course Descriptions The programme offers technical and business courses that can be fundamental or elective, according to the track chosen. For more details, please refer to the Master’s website.

the functionality of embedded systems. It discusses how the specification models can be used to make a design flow refining an initial specification down to an implementation. This is illustrated with interactive exercises in SystemC design environment.

Fundamental Courses

Networking A lot of embedded systems (e.g., PDAs, cell phones) now have graphical user interfaces, hardware and software expansion options, and network connection. This course provides an overview of embedded systems networking with a demonstration of Java networking software in laboratory sessions.

Computer Architectures The course deals with advanced computer architectures with a particular focus over the embedded systems. Topics include cache architectures, pipelined CPUs, Instruction-Level and Thread-Level Parallelism (with emphasis to superscalar architectures), DSP CPUs, and multiprocessor architectures. Software Compilers The course presents the essentials of compiler organisation, starting from front-ends and IR generation to machine dependent/independent optimization techniques. Lab sessions introduce tools for scanner and parser generation (flex, bison), portable front-ends for C (SUIF), Java bytecode interpreters etc. Real-Time Operating System and Scheduling The course gives an introduction to the basic concepts of realtime computing and focuses on the two major issues: real-time scheduling and real-time kernels. It discusses predictable scheduling algorithms and addresses the challenges and issues in the design and implementation of real-time operating systems. Specification Languages This course introduces the specification models used to describe

Digital Controls Digital controls exploit steadily increasing computational capabilities to shift emphasis from hardware to software taking full advantage of modern embedded solutions. The course illustrates the main issues of digital controls theory with the basic notions required for design and implementation of a system. Digital Signal Processing This course is aimed at illustrating the relevance and need for signal processing techniques in present-day multimedia and communications systems. It gives an introduction to multi-rate systems and filter banks, discusses optimal and adaptive filtering along with case studies (e.g., ADSL modem, smart antenna). Design Technologies This course illustrates the computer-aided methods for digital design. It discusses high-level models and languages, high-level synthesis, combinational logic synthesis for multi-level networks

Master of Science in Embedded Systems Design

(algebraic and Boolean methods, timing issues), sequential logic synthesis (FSM optimization and retiming) etc. Low Power Design This course provides an exhaustive review of state-of-the-art techniques for power estimation and optimization at different levels (e.g., system, behavioural, RTL) of designing a system. Lab sessions are organized over the usage of state-of-the-art CAD tools (e.g., synopsys design/power compiler). Validation and Verification The course introduces students to verification of hardware and software through model checking. The foundations of model checking are presented starting from the modelling of systems and properties, and then proceeding with the basic algorithms. The laboratory sessions focus on usage of model checkers. Re-programmable Systems The course deals with the high-complexity re-programmable circuits (particularly embedded systems) focusing on the automatic synthesis and generation of logic schematics from functional description in VHDL. Also, laboratory sessions introducing microcontrollers and FPGAs (e.g., Altera) are organized. Microelectronics In designing embedded systems and Systems on Chip (SoC), today one needs to be capable of integrating all the system components on a single chip. This course discusses the layout design, design of CMOS circuits, focuses on deep submicron technologies presenting possible solutions at circuit and layout levels.

167

Dependable Systems Proliferation of embedded systems forces the increased demand of dependability, timeliness, and security. The course discusses the basic concepts, measures and models used in dependable computing and communication along with testing methods, fault tolerance, diagnosis, and recovery techniques etc. Performance Evaluation The course starts with the basic concepts of queuing network models. It discusses the influence of relative sizes of different types of memories (e.g., RAM, caches, disks) on the performance of a system. Notions of parallel processing techniques and of the improvements achievable with their implementation are given. HW/SW Co-design The heterogeneity of today’s embedded systems poses new problems when it comes to specifying, simulating, designing, and optimizing such complex systems. The course discusses hw/ sw partitioning, design space exploration, interface synthesis, and estimation. Algorithms for Wireless Communication The course focuses on the design (synthesis) and performance evaluation (analysis) of algorithms. The complementary nature of mathematical analysis and computer simulation will be discussed. While the focus is on algorithms, the interaction of algorithm-architecture design will be addressed via discussion of a number of real-world commercial designs, which will serve as case studies.

168

Master of Science in Embedded Systems Design

Embedded Database The course introduces the main features of active, real-time, and main memory databases comparing them with the traditional structure of a DBMS for Information System management applications. The case of mobile, context-aware very small databases, resident on smart cards, PADs, cellular phone SIM cards or Wireless Sensors, will be thoroughly studied and an example application will be presented. Transforms Theory The course is intended as an introduction to the foundations of transform theory. The topics studied include Fourier series and integrals, Fast Fourier transform, Laplace transform, Zeta transform, and Wavelet transform. The main objective of the course is to provide a working knowledge in the subject. Statistics The course illustrates the basic principles of modelling based on electronic data, discusses the techniques of explorative analysis, continuous and discrete random variables starting from an introduction to probability theory. Also the idea of statistical induction (for both uni and multi-variate cases) is presented. Business Courses Business courses, fundamental for the business projects track, provide conceptual framework for embedded systems business. The courses are: 1. Introduction to Management, 2. Managerial Accounting and Finance,

3. Strategy, Organisation, & General Management, 4. Marketing & Purchasing, 5. Business lab. Elective Courses Here below some elective courses aim at giving students the idea about how tailoring their studies. Electives can strengthen the student’s pursuit of academic or career goals focusing on living-edge and advanced subjects. Different choices of elective courses may be proposed in different years. VLIW Architectures VLIW (Very Long Instruction Word) is a successful design philosophy for general-purpose, embedded, and DSP processors. The course points out the VLIW philosophy and discusses exception handling, speculative execution, predication, customizablity, power/energy considerations etc. Also, laboratory sessions are organized using the VEX (VLIW EXample) tool from HP labs. Retargetable Compilers The course focuses on mapping high-level language programmes written in C to embedded processors along with laboratory sessions for the same. The course presents techniques for efficient machine programmes generation including optimization for power consumption and generation of compilers for new processors (re-targetable compilers) from machine descriptions. The overall goal is to provide insight into the mutual dependencies of architectural processor features and efficient algorithm implementation.

Master of Science in Embedded Systems Design

Future Architectures The course initially describes the state-of-the-art in the design of superscalar processors and examine the main hurdles for their scalability: increasing complexity, limited parallelism of applications, memory to processor gap, power consumption, power density, soft errors, and wire delays. The course describes current research efforts in clustered microarchitectures, speculative multithreaded processors and poweraware microarchitectures, design of the memory hierarchy, data speculation techniques, and tools for programme analysis and optimization. Network-on-Chips With deep-submicron technologies, the major challenge designers of a System-on-Chip (SoC) must overcome is to provide a functionally correct, reliable operation of the interacting components. The course points out the basic concepts (e.g., synchronization), discusses buses and some state-of-the-art Networkon-Chips (NoCs). After introducing some necessary fundamental concepts, this course presents and discusses major aspects of Networks on Chip (NoC) design: integration (mapping) of functional and architectural specification, SoC communication architectures, communication trade-offs. Also, laboratory sessions are organized using a NoC simulator from NEC labs. Cryptography Security is a critical concern in some embedded systems. The course firstly gives an introduction to data security and cryptography along with cryptanalysis. Subsequently, it discusses starting from simple blockciphers to stream ciphers, Data Encryption

169

Standard (DES), Advanced Encryption Standard (AES) – Rijndael, public key cryptography (RSA), digital signatures etc. Seminars Interactive seminars offer participants tools for acquiring and improving their interpersonal skills. Thesis from Master Project The final Master’s thesis is the outcome of a research project as integral part of the study programme. This original contribution to the body of embedded systems design can be more theoretical or experimental, but always builds on a solid research effort, and on the use of appropriate concepts, methods, and tools acquired during the Master. Faculty members support and coach students during their Master’s thesis work. The thesis can be started during the third semester and completed by the end of the fourth semester. Placement ALaRI advices and supports its graduates seeking employment thanks to its network of collaborative relations with universities, research centres and industries. Such contacts aim at creating opportunity for full-time jobs, interships, or PhD positions.

170

Master of Science in Embedded Systems Design

General Information Language This programme is entirely held in English. Admission Requirements The formal requirement for admission is a Bachelor’s degree (or equivalent) from a recognised academic institution in Computer Science, Mathematics, Physics, Electronics or Electrical Engineering, Telecommunications, or other related disciplines. Good knowledge of English is required. Detailed admission requirements are available on ALaRI website at www.alari.ch/Admission/MasterOfScienceRequirements.htm Contacts Università della Svizzera italiana, USI Study Advisory Service Tel. +41 58 666 4795 [email protected]

For further information: www.alari.unisi.ch