DVB-T2 finds worldwide success

When the DVB forum published the second generation of its digital terrestrial TV standard, DVB-T2, in 2008, it did not anticipate the tremendous success this transmission system would have around the world.

The first commercial deployment of DVB-T2 was in 2009 in the UK (similar to the first-generation DVB-T, which started in the UK at late 1998) as a companion to the region-by-region analog TV switch off. The digital TV services of one multiplex were spread over the five remaining DVB-T multiplexes, and the freed multiplex was equipped with DVB-T2. This new DVB-T2 multiplex delivered up to five full HDTV programs, and the overall capacity of the UK's terrestrial digital TV network has been increased by almost 50 percent (i.e. from 108Mb/s to 160Mb/s). The designers of DVB-T2 not only doubled the efficiency of the DVB-T transmission system, but also increased the overall flexibility and reliability of terrestrial broadcast transmissions.

Another outcome of the DVB-T2 debut in the UK was the high penetration rate of the related STBs able to access both DVB-T2 and DVB-T transmissions, as well as streams encoded as MPEG-2 and MPEG-4. The price of STBs continues to fall rapidly.

2010-2011: DVB-T2 expansion

The great success of DVB-T2 in the UK paved the way for other countries to embark on the DVB-T2 adventure. (See Table 1.)

Table 1: According to the DVB Project, these countries have reported DVB-T2 acceptance as of August 2011. Deployed Finland, Italy, UK, Sweden, Zambia Trials Belarus, Denmark, Germany, Kazakhstan, Malaysia, Russia, Spain, Switzerland, Thailand Adopted Austria, Czech Republic, India, Kenya, Serbia, Singapore, Slovakia, South Africa, Sri Lanka, Ukraine The following groups report no activity, but the Southern African Development Community (SADC) has recommended DVB-T2. Angola, Botswana, Congo, Lesotho, Madagascar, Malawi, Mauritius, Mozambique, Namibia, Seychelles, Swaziland, Tanzania, Zimbabwe

After two years of deployment, the DVB-T2 system is delivering advanced digital TV services to a population of 150 million people (61 million households) in five countries. Russia, India and 30 additional countries intend to increase these figures.

A mix of advanced solutions

Contrasted with first-generation DVB-T, which used only two processing blocks (channel encoder and OFDM modulator), four functional blocks contribute to the coding and modulation of the DVB-T2 broadcast signal. (See Figure 1.)

  1. The T2-Gateway is the entry point to the T2 broadcast network: It processes multiplexes of broadcast services having a transport stream format (i.e. MPEG data stream) or a generic stream format (i.e. IP data stream) and populates the physical layer pipes (PLP), which will be distributed to the transmitter sites using the T2-Modulator Interface (T2-MI - ETSI TS 102 773).
  2. The T2 Channel Encoder is the “Robustness Generator,” which provides protection (i.e. FEC encoding) and diversity.
  3. The T2 Frame builder is the “Resources Optimizer,” which optimizes the organization of the PLPs over the transmission frames.
  4. The T2 OFDM Modulator is the “Wave Generator” building up the multicarrier waveform, equipping carriers with data and inserting necessary signals for synchronization, signalling and sounding (SSS).

Enhanced channel coding

The first generation of DVB standards addressing terrestrial broadcasts (i.e. DVB-T, -H, -SH) offer a set of transmission parameters to permit the deployment of various broadcast infrastructures targeting stationary, portable and even mobile receivers, but the level of robustness was selected for the whole broadcast network; each service inherits a common but unique attribute against transmission impairments.

In its second generation of transmission standards (i.e. DVB-S2, -T2, -C2), DVB introduced PLPs, each having a specific channel encoding process, then characterised by a specific level of robustness. While other existing broadcast systems offer two layers (i.e. ATSC-M/H) or three layers (i.e. ISDB-T) of channel encoding, DVB-T2 extends such layered protection capability to up to 256 PLPs, providing to the broadcasters the flexibility to split the transmission resource either per TV channel (i.e. each PLP carrying all components of a TV), or per population of receivers (i.e. specific level of robustness for fixed — with rooftop antenna, portable, mobile reception), or per service (i.e. HDTV, SDTV, LDTV contents being grouped in a specific PLP) etc.

Beyond its flexibility, the channel encoding process designed for T2 implements not only the state-of-the-art inner code (LDPC) and outer (BCH) codes for FEC, it also enhances their correction capabilities by a set of interleavers. The bit-interleaved coded modulation (BICM) process and the time and frequency interleavers are able to randomize the distribution of the information bits onto the T2 modulated wave.

Robustness near to the Shannon capacity limit

Robust FEC and maximization of the diversity have allowed the T2 channel encoder to exhibit “near-to-Shannon-limit” transmission performance. (See Figure 2.)

In most of the implementations, digital terrestrial networks of the first generation have been planned to deliver 24Mb/s over an area surrounding the DVB-T transmitter where the carrier-to-noise ratio is above 17dB. With DVB-T, this operating point corresponds to a 64QAM modulation and a FEC code rate of 2/3.

In the UK, where DVB-T2 upgraded an existing 17dB DVB-T network, two operating points have been considered: 256QAM, 3/5, which provides 36Mb/s at a 16dB C/N threshold; and 256QAM, 2/3, which provides 40Mb/s at an 18dB C/N threshold.

The latter was selected and increased the broadcast throughput by 66 percent without changing the radiated power in the service area. If a 64QAM, 3/5 operating point would have been chosen, the DVB-T2 network would have delivered 26Mb/s with a C/N threshold of 12dB — a smaller bit rate gain but a great +5dB gain in robustness, which will have improved the service to portable receivers as they don't benefit from the gain provided by a rooftop antenna.

Clearly, the performance gain provided by the T2 channel encoder offers broadcasters new opportunities to optimize the terrestrial broadcast platform such as reducing radiated power (lower CAPEX), increasing the number of TV programs (lower average OPEX) and/or improving indoor coverage. Furthermore, several classes of PLPs each focused on a population of receivers having specific requirements (i.e. stationary/nomadic, HDTV/SDTV capabilities, etc.) should be deployed, mixing all the above advantages.

Optimum multicarrier waveform

As for the first generation of digital terrestrial TV, DVB-T2 produces a multicarrier waveform constituting an OFDM signal in which the density of subcarrier is maximized by fixing the intersubcarrier spacing to a value equal to the inverse of the symbol duration, thus avoiding intersubcarrier interference.

Then, to provide resilience against the multipath propagation of the transmitted wave, each OFDM symbol is extended by a cyclic prefix, which fills its beginning by duplication of the finishing symbol samples. The intersymbol interference, induced by the delayed finishing samples of the previous OFDM symbol, occurs at the beginning of each OFDM symbol. This “Guard Interval,” made of sample surpluses not essential on the receive side, provides a mechanical protection against the “echoes” as long as it has a duration longer than the maximum delay of the echoes.

With DVB-T2, the set of parameters driving the COFDM waveform was enlarged:

  • New channel bandwidths for deployment from 30MHz to 3GHz (the whole VHF and UHF);
  • New FFT and Guard Interval sizes for transmission cell sizes from 3km to 300km;
  • Extended set of Pilot Patterns for adjusting the overhead from 1 percent to 10 percent. (See Table 2.)

The wide range of parameters available with T2 provides flexibility in the deployment of the system on the field (i.e. waveform and RF aspects) further complemented by the ability to provides a range of C/N thresholds (i.e. services and PLP aspects).

Wide network planning capabilities

The intrinsic capability of COFDM systems to work in the presence of echoes allows SFN operation in which echoes are actively produced by transmitters — each radiating at each instant identical digital samples — to increase the delivered power, especially to the receivers located in areas where the transmitters' coverage overlaps. (See Figure 3.)

But T2 introduced the possibility to use the multiple input single output (MISO) technique to increase signal diversity over the air. Instead of SFN transmitters radiating strictly the same samples, MISO-SFN transmitters radiate variants of such digital samples. To create two variants of the signal, an Alamouti signal coding is implemented onto pairs of adjacent subcarriers (a,b):

  • The transmitters pertaining to group A, map (a) and (b) to the adjacent subcarriers and generate a regular pattern of pilots;
  • The transmitters pertaining to the group B, map the complex conjugate (-a) and (b*) on the identical adjacent subcarriers and invert one pilot over two within the pilot pattern.

Whatever the received signal A or B, the receivers are able to demodulate, helped by the pilot pattern which identifies the type of signal “variant.”

But the real gain obtained by MISO coding is revealed when both A and B are received. In this case, the signal diversity created by Alamouti signal encoding contributes to decrease significantly the frequency selectivity of the “0dB echo” behavior of the SFN transmission channel (i.e. subcarriers issued from SFN transmitters should arrive either in-phase or in opposite phase at the receiving point, creating either a subcarrier gain or a subcarrier cancellation).

Field trials performed in Finland and Germany verified the raw power gain of 2dB to 3dB characterizing the traditional SISO-SFN operation, but revealed on top of it a nice additional gain of 2dB to 3dB with the MISO-SFN operation.

Beyond DVB-T2

As far as pure transmission performance is concerned, DVB-T2 operates near the Shannon capacity limit, and no dramatic improvement could be expected from channel encoding or modulation.

Nevertheless, additional flexibility has been considered by DVB, which produced in June 2011 an annex T2-Lite, which defined a subset of T2 parameters to allow the design of a T2 demodulator more suitable for integration in nomadic receivers.

On the network side, additional flexibility has been provided to allow mixing T2 transmission frames having different set of parameters. Thus, some T2 frames could use a 32K FFT to benefit from the larger guard interval, which maximizes the transmission cell size. Other T2 frames transmitted sequentially with the previous should use 8K FFT and offer slightly less coverage but increased intercarrier spacing — easing reception in motion.

Gerard Faria is chief technology officer at TeamCast.