Sparse Equalizers for OFDM Signals With Insufficient Cyclic Prefix

Abstract

The cyclic prefix (CP) is appended in orthogonal frequency division multiplexing (OFDM) signals to combat inter-symbol interference (ISI) and inter-carrier interference (ICI) induced by the communication channel, which limits its spectral efficiency. Therefore, inserting an insufficient CP and equalizing the resulting ICI and ISI is a method that has been circulating the literature for a while, aiming at increasing the efficiency of OFDM systems. In this paper, we propose a reduced-complexity sparse linear equalizer and a decision-feedback equalizer for OFDM signals with insufficient CP. A performance-complexity trade-off is highlighted, where we show that it is possible to equalize the received signal with a reduced complexity equalizer while having a limited performance loss. Our proposed equalizer designs are not only less complex to realize, but are shown to provide a higher data rate. The proposed equalizers are further evaluated in terms of the worst-case coherence, a metric determining the effectiveness of our used approach. Numerical results show that we can significantly and reliably reduce the order of the design complexity while performing very close to the conventional complex optimal equalizers.

Keywords

• OFDM
• sparse approximation
• linear equalizers
• decision feedback equalizers
• insufficient cyclic prefix
• worst-case coherence