Abstract:
We consider an intensity-modulated (IM) communication system in the presence of Gaussian noise. To improve the reliability of such a system, a space-only constellation-op...Show MoreMetadata
Abstract:
We consider an intensity-modulated (IM) communication system in the presence of Gaussian noise. To improve the reliability of such a system, a space-only constellation-optimal structure was proposed by Zhang under a peak and an average power constraint. However, how to utilize time resource to design signals has been an unsolved problem. In this paper, we design a complexity-efficient coded modem for IM signals under a peak- and an average-power constraint in finite-blocklength regimes. The proposed modem consists of a lattice-based multidimensional constellation bounded by the optimal shaping region, decomposable constellation mapping and demapping based on the fast Fourier transform (FFT), and a fast demodulator with near maximum-likelihood performance. To the authors' best knowledge, for the first time, the FFT and the inverse FFT (IFFT) algorithms are used for constellation mapping to reduce the involved complexity. Both theoretical analysis and Monte Carlo simulation results show that our proposed modem has significant signal-to-noise ratio gains and can support flexible-rate transmission compared to the one-dimensional optimal constellation with Gray labeling.
Published in: IEEE Transactions on Communications ( Volume: 68, Issue: 1, January 2020)