Carrier Frequency Offset Impact on Universal Filtered Multicarrier/Non-Uniform Constellations Performance: A Digital Video Broadcasting—Terrestrial, Second Generation Case Study

Abstract:

Digital terrestrial television is now widely implemented across many countries worldwide and has reached a high level of maturity. Digital Video Broadcasting–Terrestrial, Second Generation (DVB-T2) is the European standard that has been adopted or deployed in numerous European and African countries. It relies on Orthogonal Frequency-Division Multiplexing (OFDM) modulation to achieve high throughput performance. One of its main characteristics is the flexibility in the number of OFDM subcarriers, which can range from 1,024 to 32,768. In addition to rooftop antenna and portable reception, already supported by DVB-T, DVB-T2 also enables mobile reception. However, one of the major challenges associated with mobile reception is the presence of a Carrier Frequency Offset (CFO), which degrades system performance by generating Intercarrier Interference (ICI) within the DVB-T2 signal.

This paper evaluates the performance of the DVB-T2 system under the influence of CFO in the presence of both Gaussian noise and a TU6 propagation channel. Previous studies have shown that Universal Filtered Multicarrier (UFMC) and Non-Uniform Constellations (NUCs) provide better performance than conventional OFDM and Quadrature Amplitude Modulation (QAM) within the DVB-T2 framework. Therefore, this work further investigates the impact of CFO on a UFMC- and NUC-based DVB-T2 system.

The obtained results show that the performance degradation caused by CFO is significantly reduced when UFMC and NUC are employed, compared with the native DVB-T2 system. At a Bit Error Rate (BER) of 10⁻³, the CFO penalties observed in the native DVB-T2 system are 0.96 dB when only Additive White Gaussian Noise (AWGN) is considered, and 4 dB when the TU6 channel is additionally taken into account. In contrast, the corresponding penalties for the UFMC/NUC-based DVB-T2 system are reduced to 0.84 dB and 0.2 dB, respectively.