DAB: Transmitter, Receiver and SVM Classifier

Abstract

The aim of this thesis is to implement a Digital Audio Broadcast (DAB) system in base-band. In this work I designed a DAB transmitter for mode II, following the ETSI standards for DAB transmitter was designed. This mode has been chosen for simulation because of its suitability in the local area terrestrial broadcasting and to be a model that presents other transmission mode implementation. The physical modulation part of the DAB for transmission modes II as well as its receiver is implemented.
The prime focus of this thesis is on the reception side. In the receiver side, receiver synchronization has been implemented and a new classification method for the QPSK mapping proposed. Support vector Machine (SVM) is used as a QPSK Classifier and as a channel estimator for DAB system. It uses the continuous learning algorithm, for training and testing. This learning algorithm needs one-time training for the classification of first DAB symbol, and for the remaining symbols the system learns the pattern from the previous symbol estimated by classifier.
A frame-based processing was used in this study. Performance studies for AWGN and Rayleigh channels have been conducted. Bit error rate (BER) has been considered as the performance index.
The result obtained showed that the implemented system work successfully in AWGN channel. For Rayleigh fading channel, the system performance is desirable in urban area, below the speeds of 50-70 Kmph of the receiver.
All the simulation work is implemented using the Microsoft Windows Operating system and MATLAB.