On the Design of Precoder for Inter-Carrier-Interference Self Cancellation with High Spectral Efficiency

This study uses the OFDM signal, generated by general inter-carrier interference self-cancellation (ICI-SC) precoding scheme, to develop a linearly constrained OFDM subcarrier combining scheme at the receive side. The key feature of the proposed scheme is that it is able to handle the ICI caused by a large scale frequency error. Conventional ICI-SC can hardly eliminate this kind of large-frequency-error-caused ICI, because, with a large scale frequency error, a desired subcarrier in an OFDM symbol experiences ICIs mainly from its adjacent subcarriers, on which the magnitudes of these ICIs dramatically vary with frequency. The basic idea of the proposed approach is to develop a novel subcarrier combining scheme that primarily aims to eliminate the ICI from the adjacent subcarriers. On the other hand, because of their flat distribution along frequency, the ICI contributed by the subcarriers far from the desired one is handled in this study through the ICI-SC at the transmit side. In addition, we also propose a filter-bank-based approach to provide the frequency error coarse information for the design of the proposed subcarrier combining scheme. In addition, this study will develop a subspace-based algorithm to estimate the frequency error in the OFDM. We intend to construct a receive vector based on the high coherency in the samples at specific time instances of a general ICI-SC generated OFDM symbol. The constructed receive vector possesses a frequency directed structure, that allows the proposed approach to estimate the frequency error through array signal processing techniques. Before the frequency estimation, the proposed approach removes the scaling coefficient in the receive vector through a time-domain equalization process. In addition, this study will also investigate a time sample selection scheme to alleviate the potential noise enhancement issue in the equalization process.

Project ID:PB10608-2401
External Project ID:MOST106-2221-E182-011