Jari Koskelo: Channel coding of TETRA using DSP


The purpose of this thesis is to describe the channel coding algorithms used in the TETRA-system and to study the implementation of these algorithms using a digital signal processor (DSP). The focus is on block and convolutional codes.First, the basics of wireless communication systems are reviewed including different access methods. Also the properties of the mobile communication channel including fading are examined with respect to the TETRA-system. Then, the TETRA-system and the used DSP are presented. The basic principles of channel coding and the channel coding algorithms used in the TETRA-system are described.Efficient computation of the TETRA shortened cyclic block code used also in the CCITT s Recommendation X.25 is studied. Algorithms are presented for the DSP implementation of the block code using a table access and a shift register. Various representation methods of convolutional codes, including the shift register structure, the tree diagram, the trellis diagram, and the state diagram, are illustrated using the mother code of TETRA. In TETRA the convolutional coding consists of mother encoding and puncturing. The puncturing methods used are described, and in the implementation the focus is on RCPC 2/3 code (Rate-Compatible Punctured Convolutional code of rate 2/3). After studying the encoding, decoding by the Viterbi-algorithm is studied. Other channel coding algorithms used in the TETRA system are also described shortly, including interleaving, scrambling, and Reed-Muller code.The results achieved show that the computationally demanding TETRA channel decoding can be done fast enough using DSP when efficient algorithms are used. In fact, in addition to error control functions, computational capacity is available also for other receiver algorithms.

Keywords: channel coding, convolutional coding, block coding, DSP, TETRA


Updated 4.1.1996 by
anne.saarinen@hut.fi