Bit rate maximization for LP-OFDM with noisy channel estimation

The resource allocation problem to maximize the bit rate of a linear precoded orthogonal frequency division multiplexing (LP-OFDM) system is considered taking into account the effects of imperfect channel state information. A discrete bit loading algorithm is proposed, which sustains the target bit error rate even under high mean square error (MSE) of estimation. The proposed scheme enhances the robustness of the system against noisy channel estimation without significantly compromising on the system throughput. It is shown that the proposed LP-OFDM allocation is more robust to estimation noise than OFDM allocations and provides sustainable mean bit error rate performance even at high MSEs. The results are shown for a power line communication system using a well-known multipath channel model.