It has been a long road to ATSC 3.0 spanning 10 years and only coming to an end in the US after the FCC finally agreed in February 2019 to start accepting license applications from broadcasters. This comes two years after the formal standardization was completed as Seoul Broadcasting System (SBS), Korean Broadcasting System (KBS) and Munhwa Broadcasting Corporation (MBC) launched the world’s first terrestrial ATSC 3.0 broadcast network in time for the 2018 Winter Olympics.
Now the stage is set for US deployments and therefore ATSC 3.0 will loom large at NAB 2019, with a strong focus on support for interactive and hybrid services.
ATSC 3.0 accomplished roughly what DVB-T2 does in Europe when combined with the HbbTV 2.0 hybrid broadcast standard. Indeed, ATSC 3.0 has incorporated and built on many features of both, notably the use of OFDM (orthogonal frequency division multiplexing). OFDM we recall has been deployed in wide area wireless networks, including 4G/LTE as well as DVB-T2, to boost data rates. It evolved through two related developments, firstly splitting each symbol representing individual bytes of data across multiple low bit rate carriers. These carriers can then be packed closely together with overlaps, greatly increasing both throughput and resilience to inter-symbol interference.
This in turn feeds off the second development, the orthogonality, which renders each symbol totally immune from interference from any of the others. This exploits the waveform of the radio signals which can be represented by oscillating sine waves with peaks and troughs. If the multiple low bit rate carriers are separated by spaces corresponding to multiple of the wave period, then with the appropriate modulation and demodulation techniques their sum total is zero since the peaks cancel out the troughs from neighboring carriers. Effectively then they are all at right angles, or “orthogonal”. Only the actual signal carrying each symbol then has a non-zero value that can be determined through mathematical integration.
Being newer, ATSC 3.0 has naturally made some improvements on DVB-T2, with more options for the coding. It also allows use of higher QAM modulation rates for the symbols at 1024 QAM and 4096 QAM, capable of carrying high data rate 4K UHD content. It still remains to be seen how useful these options are for terrestrial transmission in the UHF band since they require high signal levels for an error-free reception.
However, ATSC 3.0 also provides very robust system variants at low bit rates using QPSK modulation. These may provide robust reception to mobile receivers possibly using LDM (Layer Division Multiplex), which is another new feature in ATSC 3.0. This is designed to make it possible to combine HD programs for rooftop reception and robust mobile reception in a single UHF channel, in an efficient way. There is also channel bonding so that content can be split across two RF channels, potentially doubling the data rate further, although twin RF tuners are then required.
The upshot then is that ATSC 3.0 has a number of features that between them support UHD services while at the same time enabling more robust transmission to mobile devices. The other key feature is support for interactivity and integration with broadband delivery, which in the case of DVB-T2 was not in the standard and relied on HbbTV.
This brings us to an interesting part of the ATSC 3.0 story because originally, out of natural desire not to waste effort reinventing already well proven technology, the ATSC planned to base its interactive content component A/344 almost entirely on the HbbTV work, adapting the specifications accordingly.
This still requires new skills to develop effectively for the TV, but those skills are now entirely related to the UI and specifics of the laid-back experience, avoiding the need to grapple unnecessarily with legacy technologies. The point here is that for laid back consumption on the couch the TV UI is still very different from say a PC UI, with usually no keyboard or mouse and a need to make navigation as simple as possible via a remote.
For this reason, ATSC has tried to make life as simple as possible by extracting features from the four most popular browsers (we presume Google Chrome, Mozilla Firefox, Opera and either Microsoft Edge or Safari). This is wrapped up in the ATSC’s A/344 standard which incorporates a subset of the current W3C web development interfaces, selecting only those available in these four major web browser stacks. Developers can therefore target TVs irrespective of which of these four browsers they have by writing to the A/344 standard and should not be too bothered at missing out on some of the less popular options. In fact, ATSC recommends adhering to A/344 even when publishing video apps just to the web since this will ensure consistency and easy portability.
The obvious next question is what interactive apps we are talking about here in the laid-back sense. Clearly some broadcasters will be looking at exploiting these A/344 features for “lean forward” apps incorporating on-screen graphics with buttons, pull downs and the usual familiar online elements. But others will be looking to enhancing lean back viewing to add value by allowing access to supplementary content, but the real killer app in the US is likely to be addressable advertising, since ATSC 3.0 for the first time enables broadcasters to substitute ads dynamically within the broadcast stream.
Here again we encounter differences between ATSC 3.0 and HbbTV. One argument ATSC makes is that the HbbTV specification does not guarantee frame accurate switching between ads, with variations in timing between devices. That can be catered for but imposes effort and overhead, for example by inserting “guard areas” around the broadcast/broadband switching points.
This comes at a time when initial commercial deployments of addressable advertising within broadcast have been confined to vertical ecosystems defined by given pay TV operators, such as Sky’s AdSmart in Europe, now being adopted by Comcast in the US after its takeover. Broadcasters, or networks, have been unable to follow because they have been denied real-time access to viewership information, as well as the ability to insert ads dynamically in individual homes. ATSC 3.0 at last will give broadcasters the tools to emulate services already proven with some pay TV operators.
This is also enabled by HbbTV in conjunction with DVB-T2, as some European broadcasters have already shown. France Télévisions demonstrated that DVB-T2 can deliver advanced advertising with targeting down to household level with HbbTV in a trial through its advertising arm France TV Publicité, in conjunction with the country’s DTT transmission company TDF along with advertisers Sofinco and St Hubert.
ATSC 3.0 has taken a similar approach in some respects, for example in the overall flow and business logic. The main difference is that with ATSC 3.0, the broadcast transmission is also IP-based and uses MPEG-DASH. This allows broadcasters to adopt mechanisms already proven for internet streaming services to insert ads into linear TV streams.
Although yet to be proven then, ATSC 3.0 does score by sharing many building blocks with OTT streaming services, allowing broadcasters and advertisers to rely on existing tools and skills from the online world. In the deficit column is the lack of backwards compatibility with legacy broadcasting infrastructure, which HbbTV does have and was considered essential at the time it was developed. ATSC decided to shed that constraint but that has resulted in delays and resistance from some broadcasters.