Iterative detection algorithms for extended MC-CDMA

Multicarrier code division multiple access (MC-CDMA) combines the advantages of CDMA and orthogonal frequency division multiplexing techniques. For time-varying channels, it can be further improved by introducing an additional spreading in the time direction (extended MC-CDMA). Two low complexity iterative detection algorithms with soft feedback are described: the iterative soft block decision feedback equalizer and a recurrent neural network based detector. Performance results are given for a typical mobile radio channel and compared to the optimum maximum likelihood detector. Introducing pilot symbol assisted channel estimation, the optimum number of pilot symbols is determined and the bit error rate performance is compared to the case of ideal channel estimation