Combining Error-correction Coding and Cutoff Rate Maximization Based Precoding

A new linear block precoding technique is combined with error-correction coding for wireless orthogonal frequencydivision multiplexing (OFDM) transmissions through frequency-selective fading channels. Our precoder is based on the maximization of the mean cutoff rate and requires only the knowledge of the average relative channel multipath gains and delays at the transmitter. Convolutional coding is used as an outer channel coding (CC). Two decoding schemes are developed for the proposed joint channel-coded and linearly-precoded OFDM communication system. The first decoding scheme consists of the cascade of maximum likelihood (ML) symbol detector and Viterbi decoder. The ML detector decodes the symbols encoded by the linear precoder, and the Viterbi decoder decodes the convolutional channel codes. The second decoding scheme is based on the iterative (turbo) decoding approach, which exploits the soft information and enables to further improve the decoding performance. Simulation results with convolutional codes for UMTS and HIPERLAN/2 channels show an improved performance of the proposed joint coding-precoding technique as compared to entirely precoding-based, entirely CCbased, and other known joint coding-precoding techniques. Among the two proposed decoding approaches, the turbo decoder provides better performance at the price of higher computational complexity and larger memory consumption.