Good error-correcting codes based on very sparse matrices
IEEE Transactions on Information Theory
| Hi-index | 0.00 |
In space division multiplexing with orthogonal frequency division multiplexing (SDM–OFDM) systems, since co-channel interference (CCI) degrades the demodulation performance, CCI suppression is essential. For CCI suppression, the turbo detector using the soft-input soft-output maximum likelihood detector (SISO–MLD) is proposed. Although SISO–MLD can deal with soft information, SISO–MLD updates only a priori information of code bits at each iteration. This prevents the performance to be improved. Meanwhile, a turbo detector with soft cancellation (SC) followed by minimum mean square error (MMSE) filter, in which the variance of the residual interference components as well as a priori information can be updated at each iteration, is well known. However, the performance of SC/MMSE detector is also limited because it is proposed to reduce the computational complexity at the cost of performance degradation. In this paper, we propose an SC-type turbo detector which uses SISO–MLD for SDM–OFDM systems. In our proposed detector, the soft cancellation unit of the SC-type turbo detector is modified to cope with the SISO–MLD. From the simulation results, the proposed SC/MLD provides a better BER performance than the turbo detector using SISO–MLD. Furthermore, the proposed SC/MLD can attain the bit error rate (BER) equivalent to that of the SC/MMSE detector with a smaller computational complexity.