Multicarrier Modulation and Coding for Multichannel Systems
Licentiatavhandling, 1997

In this thesis decision directed coherent detectors for single- and multicarrier systems employing an MMSE channel estimation technique is presented. The detectors are designed for Rayleigh fading channels and DQPSK or p/4-DQPSK modulation and exploit knowledge of the fading process correlation. Pilot symbols are not used and differential encoding is needed to resolve phase ambiguities due to decision errors. Analytical and simulated results are presented. The detectors have significantly lower bit error probability than a traditional differential detector and do not suffer from the problem of an irreducable bit error floor. Further more, a Parallel Combinatory OFDM (PC-OFDM) system is presented. The PC-OFDM system uses only a subset of the available sub-carriers in each symbol interval. A bit mapping procedure is proposed using the Johnson association scheme to choose which sub-carriers to use and M-PSK modulation is used on these sub-carriers. The PC-OFDM scheme can be designed to have a higher spectral efficiency, lower bit error probability, and lower peak-to-average power ratio than ordinary OFDM systems. Analytical and simulated results are presented for the AWGN channel as well as simulated results for the Ricean fading channel. The PC-OFDM system is however not as robust against fading as an ordinary OFDM system. The use of Rate Compatible Punctured Convolutional codes (RCPC-codes) for rate matching in multichannel systems is also presented. A multi-code DS-CDMA system is used for evaluating purposes. The RCPC-codes are shown to provide a flexible and efficient coding scheme which is superior compared to repetition encoding schemes.

multi-code DS-CDMA.

Decision directed coherent detection


Parallel combinatory signaling

rate matching


multichannel systems

Channel estimation

Multicarrier modulation


Pål Frenger

Institutionen för informationsteori, Kommunikationssystem





Technical report L - School of Electrical and Computer Engineering, Chalmers University of Technology. : 258L

Mer information