Artificial Neural NetWork for DTMF decoders

Aiello, A.; Daponte, P.; Grimaldi, D.; Michael, L.

Search

View/Open document files

110-116.pdf (609.1Kb)

Author

Aiello, A.
Daponte, P.
Grimaldi, D.
Michael, L.

Date

1999

Publisher

BPI

Show full item record

Citation

Artificial Neural NetWork for DTMF decoders / A. Aiello [and etc.] // International Conference on Neural Networks and Artificial Intelligence ICNNAI'99 = Международная конференция по нейронным сетям и искусственному интеллекту ICNNAI'99 : Proceedings, Brest, Belarus, 12–15 October 1999 / Brest Polytechnic Institute, Department of Computers and Laboratory of Artificial Neural Networks, Belarus Special Interest Group of International Neural NetWork Society, International Neural NetWork Society, Belarusian State University of Informatics and Radioelectronics (Belarus), Belarusian Academy of Sciences, Institute of Engineering Cybemetics (Belarus), Universidad Politechnica de Valencia (Spain), Institute of Computer Information Technologies (Ukraine, Ternopil) ; ed. V. Golovko. – Brest : BPI, 1999. – P. 110–116.

Abstract

In this paper, the paltem recognition characteristics of the Artificial Neural Networks (ANNs) are used to realise a real decoder for Dual Tоnе Multi Frequency signals used in the telecommunication field. A new neural architecture, the Multi Leaming Vector Quantization (MLVQ) network, is proposed It offers both greater efficiency in decoding and less sensitivity to noise. In order to solve the problem regarding input signal synchronisation, a pre-processing phase is organised. Respect of the timing parameters required by the international recommendations is assured by implementing a Finite State Machine (FSM). The prototype decoder has been realised by implementing the pre-processing phase, the MLVQ network and the FSM on the TMS320C30 Digital Signal Processor. The decoder has been tested according to the ITU-T Q.24 and Telcordia Recommendations by means of a PC-based automatic measurement station. The test results are given and compared with those obtained by a traditional decoder and by a decoder based on the Multi-layer Perceptron ANN.