Applying ATSC 3.0 NUC-2D constellations to DVB-T2: an improved performance case study

Digital Video Broadcasting-Terrestrial, second generation (DVB-T2) is a European broadcasting standard developed in 2009 to deal with the first generation Digital Video Broadcasting-Terrestrial, first generation (DVB-T) shortcomings and get closer to the Shannon limit. During the last decade, it has been the object of many researches with the purpose of an investigation of the possible improvements. While some researches are based on the FEC codes, other are focused on the optimization of the constellation shaping Quadrature Amplitude Modulation (QAM) or multicarrier modulation techniques. Non Uniform Constellations (NUCs) are the most recent constellation shaping proposed for DVB-T2 to improve its capacity. Furthermore, its is included in the American broadcasting standard ATSC 3.0, the most recent and the most efficient among all broadcasting standards. Despite the same FEC codes applied in both DVB-T2 and ATSC 3.0, the control parity check matrices used for the Low Density Parity Check (LDPC) code are different. In this paper, performance of NUCs based DVB-T2 and ATSC 3.0 are presented and compared using in first Gaussian noise only and in second Rayleigh channel. It has been proven that NUCs based DVB-T2 outperforms ATSC 3.0 by 0.78 dB at a BER of 3.10
^-3
for CR 1/2 (DVB-T2) and 13/15 (ATSC 3.0) whereas ATSC 3.0 outperforms NUCs based DVB-T2 by 0.94 dB at a BER of 3.10
^-3
for CR 3/4 (DVB-T2) and 11/15 (ATSC 3.0).