Towards a new revolutionary concept of Digital Terrestrial Television

Is it possible to optimize bandwidth and spectrum efficiency while reducing energy consumption in Digital Terrestrial Television broadcasts? An interesting work developed by the Polytechnic University of Valencia, Panasonic and Teracom, and awarded the 'Best Conference Paper Award' at IBC 2016, shows that there are alternatives...

The research work WiB, a new system concent for Digital Terrestrial Television (DTT) managed to win the coveted 'Best Conference Paper Award' at the last IBC.

The proposed system, the result of collaboration between Teracom, the Swedish DTT operator, the Universitat Politècnica de València and Panasonic Europe, represents a change in the way of understanding the reception of the DTT signal, where potentially all available frequencies in the UHF band can be used from all transmitters. The increase in interference that this entails requires the combination of a robust transmission mode, the use of discrimination by directivity of the receiving antenna and interference cancellation methods, mainly based on the concept of LDM (Layered Division Multiplexing).

WiB would allow the DTT signal to be transmitted as a higher bandwidth signal, offering a very significant reduction in the power/cost of the transmitted signal and between a 37-60% increase in capacity for the same coverage offered by the current DTT system. WiB would allow reception in mobility conditions, supporting high speeds, as well as sufficient granularity to offer local services. The authors also highlight possible more advanced developments, doubling the system capacity through MIMO, backward compatibility with current antennas or even adding an additional layer to offer mobile services. Furthermore, WiB would allow the definition of a system in which both broadcasters and mobile operators use the entire UHF spectrum, transmitting with LDM, which in the future would end the current “war” for the spectrum and would ensure, in the long term, the use of the 470-694 MHz band.

Basic principles

WiB is a new concept of Digital Terrestrial Television (DTT) developed by Teracom that represents a disruptive change with respect to conventional DTT systems by being based on a broadband signal and allowing 1-reuse of frequencies. WiB evolves from one of the latest technological advances applied to DTT technologies such as Layered Division Multiplexing (LDM).

Traditionally, a DTT network is made up of high-power transmitters that transmit high-capacity signals (around 33-40 Mbps in DVB-T2) over several channels with a bandwidth of 8 MHz (UHF band). However, this high-capacity signal is in turn more sensitive to interference, requiring adequate frequency planning (reuse-N, with N between 4-7) to distance interfering transmitters from each other. As a counterpart, and given that the required power, according to Shannon's Law, increases exponentially with capacity, high-power transmitters are necessary in this case. WiB employs a much more efficient philosophy, which translates into the potential use of all channels of the UHF band, from all transmitters, expanding the power transmitted across said channels, potentially as a single wideband signal using a single transmitter.

Reuse-1 and, for example, a QPSK modulation with ½ coding, would allow a spectral efficiency of approximately 1 bps/Hz, which is practically the same as current DVB-T2 implementations using a reuse-N (assuming N=5 and 40 Mbps in MFN or N=6 and 33 Mbps in SFN). In both cases, about 200 Mbps could be offered in the 224 MHz of spectrum assigned to DTT (470-694 MHz) that will remain after the second digital dividend in the 700 MHz band.

The simulations carried out indicate that WiB could be used with a spectral efficiency significantly higher than that 1 bps/Hz (1.37 to 1.60). Furthermore, an optimal design of the demodulation process makes it possible to exploit the fact that the interfering signal has known characteristics (same QPSK modulation). For 1 bps/Hz, with a required C/N of 0 dB, adding an interfering signal at C/I=0 dB that is treated simply as noise leaves no room for more noise, making the required C/N infinite. However, if the signal demodulation takes into account that the interference has a QPSK constellation, the required C/N ends up being 6.0 dB. Exploiting the knowledge of the interfering constellation thus allows for significant performance gains.

A commonly used DVB-T2 transmission mode today is 256-QAM 2/3 encoding (e.g. in the United Kingdom). However, with QPSK ½ coding the required transmission power (for a given coverage) is about 50 times (17 dB) lower per 8 MHz channel. The net effect is that the WiB signal would essentially require only 10% of all the power currently used (across all multiplexes) of DVB-T2.

Since the coding/modulation used in WiB is restricted, the capacity of a single UHF channel will also be restricted, in the order of 7-10 Mbps. In order to compensate for the low capacity of a UHF channel, WiB assumes that services can be distributed over several UHF channels. Assuming that the bandwidth of a basic tuner can be increased by a factor of, e.g. four, going from 8 MHz to 32 MHz this would enable a peak rate of about 28-40 Mbps at this higher bandwidth. A side effect is also greater frequency diversity, which generally has positive effects on performance if the service is appropriately interleaved within a given bandwidth (using, for example, Time Frequency Slicing).

Thanks to the ability to withstand interference from adjacent transmitters, the WiB concept allows finer granularity when delivering content, potentially using all transmitters with different content.

WiB Interference Cancellation

Through reuse-1 the receiver will experience a much lower signal-to-interference (C/I) ratio than usual for DTT, which must be seriously considered.

The first tool to withstand the highest level of interference is the use of a robust transmission mode (for example 17 dB more robust than the current DTT), which would allow a C/I close to 0 dB. The second tool (for fixed reception scenarios) is that a directional antenna generally offers very significant discrimination (attenuation) in undesired (or orthogonal polarization) directions (up to 16 dB). Finally, there are interference cancellation methods through which unwanted signals can be canceled under certain conditions, as explained below.

Suppose a signal S1 interfered with by another S2 with the same modulation, but of greater power. Interference cancellation based on LDM consists of, first, demodulating the strongest signal, re-modulating it and subtracting it from the original input signal, which is possible since the signal is perfectly known after demodulating it (assuming that this has been correct). In a final step, the desired signal S1 can be demodulated. This process is possible as long as the C/(N+I) of the signal to be demodulated (S2 followed by S1) is greater than the required C/(N+I). For a spectral efficiency of 1 bps/Hz (e.g. QPSK ½) the required C/(N+I) is close to 0 dB. The described process can be generalized and used for any number of signals as long as the C/(N+I) requirement is met for each demodulated and subtracted signal.

The figure shows an example where the powers received from transmitters TX1, TX2 and TX3 are C1, C2 and C3 respectively and with N being the noise. The weaker signal from TX3 can be demodulated if the C/(N+I) is satisfied in the process of canceling the two stronger signals and there is sufficient C/N for TX3.

For LDM-based interference cancellation to work (with reasonable complexity) in WiB all the signals involved need to be fully synchronized, so that reception resembles a traditional SFN network, but with different TXs transmitting different content. In order to enable LDM-based interference cancellation each transmitter has to insert a pilot carrier sequence that is orthogonal to all other signals received from other transmitters whose interference is to be cancelled. In general, at least three orthogonal pilot patterns are needed.

A completely different alternative for interference cancellation is to use multiple antennas that, through beam-forming techniques, can maximize the C/(N+I) of the received signal dynamically. This technique could also be combined with LDM to successively cancel the set of interfering signals.

Cost reduction in network deployment

WiB would enable a substantial reduction in capital costs (CAPEX) by reducing the power consumption required in the transmitters by nearly 90%, which would also simplify equipment. This would allow all current transmitters to be replaced by a single broadband transmitter that used 50% of the power currently used in one of them. Much of the equipment currently used in DTT transmitting stations could be dispensable and could even be located directly on the transmission towers.

The reduction in power consumption directly affects operating costs (OPEX), saving on electricity bills and increasing energy efficiency. For example, with WiB, attenuators could be eliminated in the (now unnecessary) combiners, feeders, dividers,... which currently represent losses of around 3 dB (50% power reduction).

Conclusions

All the aspects summarized here are detailed in more depth in the award-winning article, which can be consulted here. It also addresses other aspects such as the characteristics that WiB receivers should have or aspects related to the introduction of the new system replacing the current DTT networks. In addition, the authors discuss other advanced features such as the use of MIMO to increase capacity, the use of two mobile/fixed LDM layers or even spectrum sharing between a WiB-DTT signal and another WiB-mobile.

About the authors

The Aliens of this Analysis Han Sido Erik Stre (pronounc, Sweds), Jour Joan Giménez (Upv, Spain), and Peter Klenner (Panasonic, Germany).

As for the Institute of Telecommunications and Multimedia Applications (iTEAM), it is a research center integrated into the Universitat Politècnica de València (UPV) specialized in the area of ​​ICT. The Mobile Communications Group (MCG), directed by Professor Narcís Cardona, is made up of around thirty researchers who develop different activities in the area of ​​wireless communications and, particularly, around broadcasting technologies. The MCG has participated in various national and international R&D projects, highlighting its contribution to the planning and design of the DVB-T2 network in Colombia.

As a member of the DVB, ATSC and FOBTV organizations, iTEAM has actively participated in the evaluation, validation and standardization of technologies such as DVB-T/H, DVB-T2 or DVB-NGH, as well as the most recent DTT standard, ATSC3.0. On the private side, the MCG collaborates with both national and international companies in the telecommunications sector, such as Sapec (Spain), ETRI (South Korea) and NHK (Japan).

iTEAM's participation in WiB is part of the "E.T. Broadcast" project, financed by the Ministry of Economy and Competitiveness (TEC2014-56483-R) and co-financed by European FEDER funds. The project, directed by Dr. David Gómez Barquero, aims to investigate and evaluate innovative technical solutions for the next generation of DTT systems, such as MIMO (Multiple-Input Multiple-Output) systems, frequency aggregation techniques (time-frequency slicing and channel bonding), Layered-Division Multiplexing (LDM), and non-uniform constellations, in terms of increased capacity, coverage and spectral efficiency; and the design of low-complexity signal processing algorithms for receivers with a reduced impact on system performance.