The performance of the united demodulator of MLSE and Viterbi decoder with interleaving (UDMVI)

Abstract

Many techniques are introduced for mobile communication. In order to compensate the effect of multipath fading and improve the BER performance, equalization and forward error correction (FEC) are used. MLSE is a kind of equalizer and it shows good performance. The combination of a convolutional coder and a Viterbi decoder with soft decision is a kind of FEC and it also shows good performance. However, they conventionally work separately. We proposed UDMV (united demodulator of MLSE and Viterbi decoder). The states of the Viterbi algorithm for UDMV is given as a sequence of information bits, while that for MLSE is given as a sequence of coded data. We also applied interleaving to UDMV. It is called UDMVI (UDMV with interleaving). UDMVI can improve 1.0 dB of Eb/No at a BER of 0.1% compared with CMSDI (MLSE and soft decision Viterbi decoder with interleaving). Furthermore, we improved the performance of UDMVI under the condition that no delay paths exist using a transmission technique. We called the technique ADPG (adaptive delay path generation). When ADPG is introduced, UDMVI can show its ability exhaustively in all conditions. ADPG with 2 branches improves 2.5 dB and ADPG with 3 branches improves 3.0 dB of Eb/No at a BER of 0.1% on I path static channel respectively.