Optimal and efficient bit loading for OFDM in the presence of channel uncertainty

In practice only imperfect channel state information (CSI) is available at the transmitter due to noisy and time-varying channel. In this paper, we consider optimal bit loading for orthogonal frequency division multiplexing (OFDM) with imperfect CSI subject to channel estimation. The frame length comes in as an important control parameter, while soft channel estimation and the Cramer-Rao bound are used to include the effect of channel estimation. We modify the conventional water-filling to take the maximal allowed rate on each subcarrier into account while considering CSI error. Its continuous output rates are optimally quantized by a non-iterative algorithm. Simulations show that the frame length is an important system parameter and that the computing time for the proposed algorithm linearly increases with the increasing number of subcarriers N.