Belépés címtáras azonosítással

magyar nyelvű adatlap

angol nyelvű adatlap

Information Processing Laboratory Exercises

A tantárgy neve magyarul / Name of the subject in Hungarian: Információfeldolgozás laboratórium

Last updated: 2008. augusztus 27.

Electrical Engineering

Specialization Embedded Information Systems

M.Sc. program

A fenti forma a Neptun sajátja, ezen technikai okokból nem változtattunk.

A kötelező előtanulmányi rendek grafikus formában itt láthatók.

The aim of the measurements is to learn in detail some information processing algorithms and their software tools frequently used in embedded systems. During the measurements the students utilize the elementary signal processing tools (e.g. averaging, filtering, discrete Fourier transform), but the aim is the development and investigation of complex systems. The subject consists of 5 measurements, each one is 8 hours long. The measurements are based on signal processing boards (equipped by Analog Devices DSPs), and the „mitmót”, the modular microcontroller-based platform developed at the Department of Measurement and Information Systems. Most of the exercises are based on real physical systems or their model. The software background is provided by LabView, Matlab, and the Visual DSP++ development system.

Use of the LabView program, steps of the development of virtual instruments. Simple exercises: timing, signal generation, displaying. Implementation of a virtual instrument (chosen from a list). Possible instruments: function generator, spectrum analyzer, oscilloscope, equalizer. The development is supported by built-in functions.

Learning of the VI set given by the hardware, steps of the development of a new project. Simple exercises: thermometer, reaction time meter. Implementation of an embedded system (chosen from a list). Possible systems: temperature control, remote control of a toy-car, sensor network for data acquisition. The development is supported by built-in functions.

Implementation of the LMS algorithm. Versions of the LMS algorithm, the XLMS algorithm. Investigation of adaptive transversal (FIR) filters. Identification by the LMS algorithm. Adaptive echo cancelation in electronic and acoustic systems.

Implementation of a classification system by multi-level processing. Processing of vibration and sound signals: extraction of the main parameters by time domain and frequency domain methods, classification by neural and fuzzy systems. Investigation of parameter setting and learning of neural networks. Investigation of parameter setting of fuzzy systems. Musical sound recognition by neural and fuzzy systems.

Signal transmission on radio channel. Implementation of synchronization of sampling. Use of interpolation techniques. Sampling of acoustic signals by “mitmót”, fusion on DSP. Better exploitation of the bandwidth: compression techniques. Influence of the number of the sensors (the number of the sensors is equal or greater than or less than required). Feedback in sensor networks.

· Attendance on the exercises is obligatory.

· Each student group has to write a measurement report on each measurement. The reports are evaluated by marks. Failed measurements are to be repeated.

· The final mark is the average of the marks of the measurement reports. Rounding is up to the next integer from 0.50.

At most 2 exercises can be additionally accomplished, independently from the origin of the failure. In case of more failed exercises (e.g. serious illness), accomplishment is to be discussed with the owner of the subject.

László Sujbert (owner of the subject), Tamás Dabóczi, Csaba Tóth