WASHINGTON—Immersive sound. Targeted, device-waking emergency alert functions. Hybrid broadcast-broadband functionality. All this, and enhanced video, cached on-demand delivery, personalized service and configurations not yet imagined.
This is the promise of ATSC 3.0, the broadcast media transmission standard in the works at the Advanced Television Systems Committee, a group of engineers volunteering their time to invent a new media platform for the masses that combines the best of broadcasting and the Internet, and steps it up a notch. They held their annual meeting in Washington, D.C. this week to report on their progress.
The standard is on track to be completed and ready to roll around a year from now. One of the key features is extensibility. It is designed to accommodate the unknown—defined in the documents as “generic” data.
“‘Generic’ also refers to something not yet invented,” said Dr. Richard Chernock, chief science officer of Triveni Digital and the chairman of ATSC Technology Group 3, the group in charge of developing the standard. “It’s ‘extensible’ in that we don’t need to know the structure, but we have created the capacity to carry it.”
After opening remarks, Dr. Chernock kicked off the two-day meeting with an overview of 3.0, a construct consisting of five layers. The foundation is a systems discovery and signaling layer known as the “bootstrap,” which received full approval as a standard in March. (See “First Element of ATSC 3.0 Approved for Standard,” March 28, 2016.) The second layer comprises transmission using orthogonal frequency-division multiplexing, or OFDM (versus the eight vestigial sideband transmission used in today’s U.S. broadcast television standard.) No. 3 is the protocol layer that defines TV content as data, or IP files. (Layers No. 2 and 3 together comprise the “physical layer.”) No. 4 is a presentation layer that provides for 4K, UHD, high- and standard-definition multicasting and immersive, object-based audio. No. 5 is the software applications layer that renders a web-like experience through the TV and secondary ancillary screens.
Mark Earnshaw of Coherent Logix presented a detailed tutorial on the bootstrap and physical layers. The bootstrap element of the standard gets the transmitted files into the radio frequency waveform. A “frame” is the largest physical layer container for carrying data traffic he said. Each contains one or more subframes.
“The bootstrap serves as a robust universal entry point into a waveform,” he said. “It contains a small amount of controlling information that enables the receiver to decode the first part of the preamble,” or the next component in a frame. “The preamble is divided into L1-Basic and L1-Detail. L1-Basic has a fixed length of 200 bits and signals parameters that enable the receiver to decode L1-Details, and the initial processing of the first frame. L1-Detail has a variable length as required and signal parameters describing the remaining subframes. It also contains parameters required to extract and decode PLPs [physical layer pipes], which carry the actual payload.”
The bootstrap further contains signaling bits that define what’s coming down the pipe and is designed for the addition of further signaling bits for the generic data mentioned previously.
Earnshaw also explained that ATSC 3.0 has a wider operating capacity than ATSC 1.0, the current standard, and is flexible in that a lower capacity is more robust, and vice versa. He also noted how the standard’s use of OFDM accommodates single frequency networks, which in turn can increase a signal coverage area throughout problematic terrain using a multiple-input, single-output, or MISO, transmission scheme. Also, multiple-input, multiple-output, or MIMO, technology can be used to increase channel capacity, though multiple antennas at the receiver are required. Channel bonding is another way to increase channel capacity, he said.
Jae-Young Lee of ETRI provided an update on the “Plug Fests” in which various vendors have come together to “confirm a common understanding of 3.0 specs, specifically, A/321 and A/322,” the bootstrap and physical layers. A total of 150 tests were done at the first Plug Fest in Shanghai last November, and 212 tests were conducted at the second Plug Fest in Baltimore in March. “Several implementation/understanding problems were identified,” Lee said.
Merrill Weiss gave a presentation on studio transmitter links, which connect the transport layer to the physical layer, and also transport pre-processed physical layer data to the transmitters. “The thing that has to be focused on is the precise emission time of the bootstrap,” he said. “The bootstrap must be emitted at a very precise time, down to the nanosecond.
Weiss said precise emission time and frequency control for the transmitter was necessary, and must back-time physical frame construction, to transmit the bootstrap from the antenna at a specified instant, and therefore buffering is necessary. He also mentioned the three STL types—microwave, satellite and fiber, and said that satellite “is particularly useful for single frequency networks.”
The other factor to consider is the reliability of the link, he said. “It must compensate for signal fading on paths.”
He also talked about using a configuration manager that sits between a scheduler and a system manager. “Say I want 12 PLPs each with a given level of robustness, and a frame length of one second. That configuration manager proposes a detailed optimization of the system that can be tweaked. This is not something you want to do manually for an optimum solution. I may want different configurations for different dayparts. These can be stored somewhere. We can’t tell equipment suppliers exactly how to do this, but we can give them the framework to do so. The system manager can be set to automatically go to pre-set No. 3, for example, depending on the services carried.”
Weiss said the highest bit rate would be around 157 Mbps, so STLs must be designed to accommodate that. However, as a practical matter, the signal won’t be maxed out all the time, so 60 Mbps is more likely, he said. “You try to make the signaling always a bit more robust than your data.”
(TV Technology will present further coverage of the 2016 ATSC meeting next week. Also, the author welcomes corrections and clarifications from the contributors herein cited.)