A Comparative Study of Unipolar OFDM Schemes in Gaussian Optical Intensity Channel

We study the information rates of unipolar orthogonal frequency division multiplexing (OFDM) in discrete-time optical intensity channels (OIC) with Gaussian noise under average optical power constraint. Several single-, double-, and multicomponent unipolar OFDM schemes are considered under the assumption that independent and identically distributed (i.i.d.) Gaussian or complex Gaussian codebook ensemble and nearest neighbor decoding (minimum Euclidean distance decoding) are used. We obtain an array of information rate result. These results validate existing signal-to-noise-and-distortion-ratio (SNDR) based rate analysis, establish the equivalence of information rates of certain schemes, and demonstrate the evident benefits of using component-multiplexing at high signal-to-noise-ratio (SNR). For doubleand multi-component schemes, the component power allocation strategies that maximize the information rates are investigated. In particular, by utilizing a power allocation strategy, we prove that several multi-component schemes approach the high SNR capacity of the discrete-time Gaussian OIC under average power constraint to within 0.07 bits.