Optical OFDM Waveform Construction by Combining Real and Imaginary Parts of IDFT

Abstract

In optical communication systems, orthogonal frequency division multiplexing (OFDM) is widely used to combat inter-symbol interference (ISI) caused by multipath propagation. Optical systems which use intensity modulation and direct detection (IM/DD) can only transmit real valued symbols, but the inverse discrete Fourier transform (IDFT) or its computationally efficient form inverse-fast Fourier transform (IFFT) required for the OFDM waveform construction produces complex values. Hermitian symmetry is often used to obtain real valued symbols. For this purpose, some trigonometric transformations such as discrete cosine transform (DCT) are also used, however these transformations can eliminate the ISI only under certain conditions. In this paper, we propose a completely different method for the construction of OFDM waveform with IFFT to obtain real valued symbols by combining the real and imaginary parts (CRIP) of IFFT output electrically (E-CRIP) or optically (O-CRIP). Analytical analysis and simulation works are presented to show that compared to the Hermitian symmetric system, the proposed method slightly increases the spectral efficiency, eliminates ISI, significantly reduces the amount of needed calculation and does not effect the error performance. In addition, the O-CRIP method is less affected by clipping noise that may occur due to the imperfections of the transmitter front-ends.

A DCT/DST Based Fast OFDM Method in IM/DD Systems

“…2. This method combines the DCT and DST to create new pre-coding matrices and utilizes the method proposed in [23] instead of the conventional FFT-based OFDM structure with Hermitian symmetry. Basically, to produce real-valued symbols on the transmitter side in this method, the corresponding real and imaginary components of the complex-valued symbols generated at the output of the IFFT are essentially summed instead of employing Hermitian symmetry at the input of the IFFT.…”

“…The study reveals around a 3 dB decrease in PAPR for both methods without a loss in BER performance. This paper introduces a novel pre-coding method which combines the DCT and discrete sine transform (DST) to create new pre-coding matrices and utilizes the method defined in [23]. This approach offers all the advantages of pre-coding PAPR reduction techniques while ensuring efficient PAPR reduction without sacrificing BER performance.…”